Non-alcoholic Non-alcoholic Fatty Liver Fatty Liver

Disease in PCOSDisease in PCOSWalter Futterweit, M.D., F.A.C.P., Walter Futterweit, M.D., F.A.C.P.,

FACEFACEClinical Professor of MedicineClinical Professor of Medicine

Division of EndocrinologyDivision of EndocrinologyMount Sinai School of MedicineMount Sinai School of Medicine

Liver Histology: the Liver Histology: the HepatocyteHepatocyte

Arranged into Arranged into cords which are cords which are separated by separated by vascular sinusoidsvascular sinusoids

Mostly bright pink Mostly bright pink hepatocyte hepatocyte cytoplasmcytoplasm

They are separated They are separated by the clear space by the clear space of sinusoidsof sinusoids

Non-alcoholic Fatty Liver Non-alcoholic Fatty Liver Disease is Common- Disease is Common- NAFLD NAFLD

&& NASHNASH pathology pathology Population PrevalencePopulation Prevalence:: NAFLD 16-23% NAFLD 16-23% NASH 2-6%NASH 2-6% (varies by ethnicity, BMI, age(varies by ethnicity, BMI, age and gender globally)and gender globally) NASHNASH has more ballooning of has more ballooning ofhepatocytes->PMNeutrophylichepatocytes->PMNeutrophylicand LC infiltrates, cellular lobularand LC infiltrates, cellular lobularnecrosis, Mallory bodies, some fibrosisnecrosis, Mallory bodies, some fibrosis

1) Ballentani S, et al. Ann Intern Med 2000;132:112. 1) Ballentani S, et al. Ann Intern Med 2000;132:112.

2) Clark J, et al. Gastroenterology 2002; 122:1649.2) Clark J, et al. Gastroenterology 2002; 122:1649.

Increasing Severity of NAFLD-Increasing Severity of NAFLD->NASH>NASH

Hepatocyte ballooning with hyaline Mallory bodies. Pericullar and perisinusoidal fibrosis &collagen

Spectrum of NAFLDSpectrum of NAFLD Simple hepatic steatosis (NAFLD) is most Simple hepatic steatosis (NAFLD) is most

frequent, usually benign, and asymptomaticfrequent, usually benign, and asymptomatic

10% have or develop liver injury and 10% have or develop liver injury and necroinflammation which is nonalcoholic necroinflammation which is nonalcoholic steatohepatitis (steatohepatitis (NASHNASH) and up to 20% of ) and up to 20% of NASH subjects may progress to more subjects may progress to more advanced liver diseaseadvanced liver disease

Cirrhosis may occur in Cirrhosis may occur in ~~ 30% of 30% of NASH patients and some may develop patients and some may develop hepatocellular carcinoma (increased 2-3 fold hepatocellular carcinoma (increased 2-3 fold in T2DM)in T2DM)

Browning JD et al. Hepatology Browning JD et al. Hepatology 2004;40:1387 2004;40:1387 Powell EA et al. Ann Int Med Powell EA et al. Ann Int Med 2005;143:7532005;143:753

Spectrum of NAFLD Spectrum of NAFLD (cont(cont’’d)d)

The increased incidence of global obesity is The increased incidence of global obesity is distinctly a NAFLD contributing factor distinctly a NAFLD contributing factor associated with a higher IR & MS prevalenceassociated with a higher IR & MS prevalence

A large 2,287 multi-ethnic population-A large 2,287 multi-ethnic population-based study in the USA found a 34% based study in the USA found a 34% prevalence of hepatic steatosis. Nl. ALT in prevalence of hepatic steatosis. Nl. ALT in 79%. Caucasian males had almost 2-fold 79%. Caucasian males had almost 2-fold increase of NAFLD than women (42% vs increase of NAFLD than women (42% vs 24% in women). Highest incidence was in 24% in women). Highest incidence was in Hispanic obese IR women with hepatic Hispanic obese IR women with hepatic steatosis in 45%. Lowest incidence was in steatosis in 45%. Lowest incidence was in African Americans (24%) and higher in African Americans (24%) and higher in womenwomen

Browning JD et al. Hepatology Browning JD et al. Hepatology 2004;40:1387 2004;40:1387 Powell EA et al. Ann Int Med Powell EA et al. Ann Int Med 2005;143:7532005;143:753

NAFLD: DX and ImagingNAFLD: DX and Imaging Liver biopsy is the Liver biopsy is the ““gold standardgold standard”” for diagnosis (?when for diagnosis (?when

to Bx)to Bx) US: sensitivity 60-99%US: sensitivity 60-99% specificity 84-95%specificity 84-95%

100% sensitive in detecting >33% fat on biopsy 100% sensitive in detecting >33% fat on biopsy Imaging generally underestimates presence of steatosisImaging generally underestimates presence of steatosis

No imaging modality allows staging of NAFLDNo imaging modality allows staging of NAFLD CT scanning reveals low density hepatic parenchyma CT scanning reveals low density hepatic parenchyma

with occasional focal areas which may be misread as with occasional focal areas which may be misread as masses. MRI can differentiate masses and may allow a masses. MRI can differentiate masses and may allow a more quantitative assessment of fatty infiltration of livermore quantitative assessment of fatty infiltration of liver

No reliable serum markers have been found to stage No reliable serum markers have been found to stage NAFLDNAFLD

Joy D, et al. Eur J Gastroenterol Hepatol 2003; 539. Saadeh S, Younossi Z, et al. Gastroenterology 2002;123:745

Scatarige JC et al. J Ultrasound Med 1984;3:914.

Link of NAFLD to Metabolic Link of NAFLD to Metabolic SyndromeSyndrome

NAFLDNAFLD 70% of obese population70% of obese population USA prevalence 23-34%USA prevalence 23-34% Risk factors similar to Risk factors similar to

those of NASH & central those of NASH & central obesity obesity

Obesity predisposes to the Obesity predisposes to the frequent presence of MS-> frequent presence of MS-> ~~50% have IGT or T2DM, 50% have IGT or T2DM, and/or dyslipidemia (TG)and/or dyslipidemia (TG)

Nondiabetics with MS and Nondiabetics with MS and increased data suggestive increased data suggestive IR have a prevalence of IR have a prevalence of NAFLD NAFLD >> those with lower those with lower IR indicesIR indices

Ludwig J et al. Mayo Clin Proc Ludwig J et al. Mayo Clin Proc 1980;55:4431980;55:443

Angelico F et al. JCEM 2006;90:1578Angelico F et al. JCEM 2006;90:1578

NASHNASH Risk factors: obesity, type 2 Risk factors: obesity, type 2

DM, hyperlipidemia (mostly DM, hyperlipidemia (mostly hypertriglyceridemia): the hypertriglyceridemia): the MSMS

Treatment with TZD may Treatment with TZD may reduce hepatic acute phase reduce hepatic acute phase reactants reactants -> -> improved improved insulin sensitivity & insulin sensitivity & inflammation. ? Metformin inflammation. ? Metformin helpful as well. Confirmatory helpful as well. Confirmatory tests of liver pathology not tests of liver pathology not done as yet with TZDs or done as yet with TZDs or metforminmetformin

10-11% wgt loss and exercise 10-11% wgt loss and exercise may reverse NASH. Biopsy may reverse NASH. Biopsy Dx pre and post-biopsy Dx pre and post-biopsy reported in PCOS case reportreported in PCOS case report

Brown AJ et al. Endocr Pract 2005;11:319Brown AJ et al. Endocr Pract 2005;11:319 Bugianesi E et al. Hepatology 2005;42:987Bugianesi E et al. Hepatology 2005;42:987 Albano E et al. Aliment Pharmacol Ther Albano E et al. Aliment Pharmacol Ther

2005;22:Suppl 2:712005;22:Suppl 2:71 Zhou G et al. J Clin Invest 2001;108:1167Zhou G et al. J Clin Invest 2001;108:1167

Non-alcoholic Fatty Liver Non-alcoholic Fatty Liver Disease and the Metabolic Disease and the Metabolic

SyndromeSyndrome

NAFLD: Obesity 60-95%NAFLD: Obesity 60-95% DM 21-55%DM 21-55% hTG 20-92% hTG 20-92% Morbidly obese: >95% NAFLDMorbidly obese: >95% NAFLD 25% NASH (biopsy25% NASH (biopsy))

Matteoni C, et al. Gastroenterology 1999;116:1413Matteoni C, et al. Gastroenterology 1999;116:1413 Dixon J, et al. Gastroenterology 2001; 121:91 Dixon J, et al. Gastroenterology 2001; 121:91

Angulo P, et al. Hepatology 1999; 30:1356. Angulo P, et al. Hepatology 1999; 30:1356. and NEJM 2002;346:1221and NEJM 2002;346:1221

NAFLD: The Hepatic Component of NAFLD: The Hepatic Component of

the Metabolic Syndromethe Metabolic Syndrome?? Association with biochemical manifestations of IR Association with biochemical manifestations of IR

Is NAFLD a predictor of MS in general population? Is NAFLD a predictor of MS in general population? In a study of 326 subjects -> 29% NAFLD, 13% MS. In a study of 326 subjects -> 29% NAFLD, 13% MS. Those with the MS had an NAFLD prevalence of 71%. Those with the MS had an NAFLD prevalence of 71%. and also had significantly higher glucose, insulin, and also had significantly higher glucose, insulin, TG, serum ALT, BMI and TG, serum ALT, BMI and central obesitycentral obesity than non- than non-NAFLD subjectsNAFLD subjects

Pts with T2DM and NAFLD have a 2-3 fold higher Pts with T2DM and NAFLD have a 2-3 fold higher incidence of acute MI than those without NAFLD incidence of acute MI than those without NAFLD independent of BMI. NAFLD likely marker of CHD ?independent of BMI. NAFLD likely marker of CHD ?

Zeiber-Zagi S et al. Hepatology Zeiber-Zagi S et al. Hepatology 2005;42:617A (Abstract#1069)2005;42:617A (Abstract#1069) Kessler A et al. Hepatology 2005;42:623A Kessler A et al. Hepatology 2005;42:623A (Abstract#1084) ( from the 56th Annual (Abstract#1084) ( from the 56th Annual Meeting of Am Assoc for Study Liver Meeting of Am Assoc for Study Liver Diseases), Kim HJ et al. Arch Int Med Diseases), Kim HJ et al. Arch Int Med 2004;164:2169.2004;164:2169.

NAFLD: The Hepatic Component of NAFLD: The Hepatic Component of

the Metabolic Syndromethe Metabolic Syndrome? ? (cont(cont’’d)d)

NAFLD in lean and overweight Korean NAFLD in lean and overweight Korean population demonstrated 16% and 23% population demonstrated 16% and 23% prevalence, respectively (increased CHO prevalence, respectively (increased CHO intake). NAFLD was associated with IR intake). NAFLD was associated with IR in both the lean and overweight subjectsin both the lean and overweight subjects

Central obesity was signif correlated Central obesity was signif correlated with NAFLD in the lean. Its association with NAFLD in the lean. Its association with MS variables were similar to with MS variables were similar to overweight without NAFLDoverweight without NAFLD

Zeiber-Zagi S et al. Hepatology 2005;42:617A Zeiber-Zagi S et al. Hepatology 2005;42:617A (Abstract#1069)(Abstract#1069) Kessler A et al. Hepatology 2005;42:623A Kessler A et al. Hepatology 2005;42:623A (Abstract#1084) ( from the 56th Annual Meeting of (Abstract#1084) ( from the 56th Annual Meeting of Am Assoc for Study Liver Diseases), Kim HJ et al. Arch Am Assoc for Study Liver Diseases), Kim HJ et al. Arch Int Med 2004;164:2169.Int Med 2004;164:2169.

NAFLD and Insulin NAFLD and Insulin ResistanceResistance

Frequently present in peripheral IR such as T2DM Frequently present in peripheral IR such as T2DM and obesity. Severity NAFLD and obesity. Severity NAFLD ~~ increased IR in increased IR in nondiabeticsnondiabetics

Hepatic IR likely plays a role in the pathogenesis and Hepatic IR likely plays a role in the pathogenesis and progression of NAFLD, and increases with age (Mprogression of NAFLD, and increases with age (M~~F)F)

Excessive triglyceride accumulation in liver Excessive triglyceride accumulation in liver IR association with increased hepatic iron levelsIR association with increased hepatic iron levels Decreased plasma adenopectin (exclusion of T2DM) is Decreased plasma adenopectin (exclusion of T2DM) is

associated menassociated men>>women with increased fat content women with increased fat content and necroinflammation in NAFLD and mayand necroinflammation in NAFLD and may->->NASH. NASH. These may be related to reduced hepatic insulin These may be related to reduced hepatic insulin sensitivitysensitivity

Bugianese E et al. JCEM 90:3498 Targher G et al 2004;Clin Endocrinol Bugianese E et al. JCEM 90:3498 Targher G et al 2004;Clin Endocrinol

61:70061:700

Multiple Multiple ““HitsHits”” to NASH to NASH

Insulin resistanceInsulin resistance

FFA synthesisFFA synthesis FFA delivery to FFA delivery to

liverliver FFA degradationFFA degradation TG release from TG release from

liverliver TG in TG in

hepatocyteshepatocytes

Normal liver

Simple steatosis

NASHCirrhosis

HCC

Oxidative stressCytokinesFFAs, low adenopectinIncreased Iron

Leptin? Increased-> fibrosis ?TNF-alpha cytokine

Adapted from Collantes R, Ong JP, and Younossi ZM, Cleveland Clinic Journal of Medicine 2004; 71(8):657.

Global Prevalence of Global Prevalence of NAFLD: USNAFLD: US

3%

14%

7%

16% 16%

34%

76%

0%

10%

20%

30%

40%

50%

60%

70%

80%

China (1) India (2) Japan(3) Italy (4) Korea (5) Korea (5) Italy (4)

LEANLEAN

OVTOVT

OBSOBSALLALL

1) Shen L, et al. World J of Gastroenterology 2003; 9:1106.2) Singh S, et al. Tropical Gastroenterology 2004; 25:76.3) Omagari Ket al. J of Gastroenterology and Hepatology 2002; 17:1098.4) Bellentani S, et al.. Ann Int Medicine 2000; 132:112.5) Kim H, et al. Arch Int Medicine 2004; 164:2169.

PCOS and the MSPCOS and the MS PCOS is the most common cause of the MS and PCOS is the most common cause of the MS and

related disorders in young womenrelated disorders in young women

Prevalence is about 7-fold higher than that of a Prevalence is about 7-fold higher than that of a female population of similar age in the USAfemale population of similar age in the USA

Changes in BMI in many countries compared to Changes in BMI in many countries compared to those in the USA reduces its prevalence ~ 4-fold those in the USA reduces its prevalence ~ 4-fold that of a young female population of similar agethat of a young female population of similar age

Apronidze T et al. J Clin Endocrinol Metab 2005;90:1929Apronidze T et al. J Clin Endocrinol Metab 2005;90:1929

Carmina E. J. Hepatology 2007;47:313Carmina E. J. Hepatology 2007;47:313

Age and the Metabolic Syndrome Age and the Metabolic Syndrome Apridonidze. Apridonidze. JCEM 2005JCEM 2005

AIMS of Study of NAFLD in AIMS of Study of NAFLD in PCOSPCOS

Prevalence of NAFLD inPrevalence of NAFLD in

women with PCOS by USwomen with PCOS by US

criteriacriteria

Predictors of NAFLD in women Predictors of NAFLD in women with PCOSwith PCOS

Gambarin-Gelwan SV et al Clin Gastroenterol Hepatol Gambarin-Gelwan SV et al Clin Gastroenterol Hepatol 2007;5:4962007;5:496

METHODSMETHODS Retrospective study Retrospective study Study population: Study population: • • 88 consecutive pre-menopausal 88 consecutive pre-menopausal

women with PCOSwomen with PCOS from 4/04 to 11/04 (office setting)from 4/04 to 11/04 (office setting) •• ≥≥18 yo (median 31.4yrs)18 yo (median 31.4yrs) •• PCOS: menstrual dysfunction, PCOS: menstrual dysfunction, hyperandrogenism, and the hyperandrogenism, and the exclusion of other causes (1990 NIHC)exclusion of other causes (1990 NIHC) Exclusion criteria: (PCOS women with regular Exclusion criteria: (PCOS women with regular

menses)menses) •• Heavy alcohol use (Heavy alcohol use (>>1-2 drinks, or 1-2 drinks, or >>20g 20g

EtOH/day)EtOH/day) •• Known liver diseaseKnown liver disease

Zawadski, JK, Dunaif, A. Current Issues in Endocrinology and Metabolism, Blackwell Scientific Publications, Boston 1992. p.377.

METHODS: DataMETHODS: Data

Clinical historyClinical history Blood pressureBlood pressure Height and weightHeight and weight Fasting glucose and insulin Fasting glucose and insulin Fasting lipid profile Fasting lipid profile 75g oral glucose tolerance test75g oral glucose tolerance test Liver chemistriesLiver chemistries Abdominal and Pelvic USAbdominal and Pelvic US

METHODSMETHODS:: Data Data Body mass index (BMI) :Body mass index (BMI) :

““LeanLean”” BMI < 25 kg/m BMI < 25 kg/m²² ““OverweightOverweight”” BMI BMI 25 but < 30 kg/m 25 but < 30 kg/m²² ““ObeseObese”” BMI BMI 30 kg/m 30 kg/m²² HOMA-IR (fasting insulin x fasting glucose / 405)HOMA-IR (fasting insulin x fasting glucose / 405) homeostasis model assessment: a surrogate insulin homeostasis model assessment: a surrogate insulin resistance measureresistance measure Severity of hepatic steatosis (US): graded asSeverity of hepatic steatosis (US): graded as nonenone mild mild moderatemoderate severesevere

n Saadeh S, Younossi Z, et al. Saadeh S, Younossi Z, et al. GastroenterologyGastroenterology 2002; 123:745 2002; 123:745..n Joseph A, et al. Joseph A, et al. Br J RadiolBr J Radiol 1979;57:23. 1979;57:23.

RESULTS: % Women on PCOS RESULTS: % Women on PCOS TherapyTherapy

0%

5%

10%

15%

20%

25%

30%

35%

40%

Perc

en

t o

f S

ub

jects

36%

27%

11%

25%

Neither Metformin OCP Both

a

c d

b

a: ABSENT: the echogenicity of the liver parenchyma is slightly greater or equal to that of the renal cortex; clear visualization of the diaphragm and intrahepatic vessels

b: MILD: slight diffuse increase in fine echoes in the liver parenchyma with normal visualization of the diaphragm and intrahepatic vessel borders

c: MODERATE: moderate diffuse increase in fine echoes in the liver parenchyma with slightly impaired visualization of intrahepatic vessels and diaphragm

d: SEVERE: marked increase in fine echoes with poor or non-visualization of the intrahepatic vessel borders, diaphragm and posterior or deep portion of the right lobe of the liver.

Courtesy of Hsu-Chong Yeh, MD (MSSM)

Grading of Hepatic Grading of Hepatic SteatosisSteatosis

RESULTS:RESULTS: PCOSPCOS Subjects by BMI Subjects by BMI

GroupGroup

0%

20%

40%

60%

80%

100%

Per

cen

t

Obese

Overweight

Lean

N=88

RESULTSRESULTS: Prevalence of : Prevalence of Steatosis by BMI GroupSteatosis by BMI Group

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Lean Overweight Obese Overall

Per

cen

t

Mild Moderate Severe

39%

54%

70%*

55%

*p<0.01 vs. leanpopulation estimates

RESULTSRESULTS

Prevalence of Steatosis by HOMA-IR Quartile

0%

20%

40%

60%

80%

100%

I II III IV

Per

cen

t

Mild Moderate Severe

37%50%

60%70%*

*p<0.05 vs. Quartile I

VariableVariable All All SubjectsSubjects

Steatosis Steatosis AbsentAbsent

Steatosis Steatosis PresentPresent

P-P-valuvalu

ee

Age, yearsAge, years 31.4 31.4 (N=88)(N=88)

29.0 29.0 (N=40)(N=40)

32.0 32.0 (N=48)(N=48)

0.280.2800

EthnicityEthnicity-Ashkenazi Jewish-Ashkenazi Jewish-Other White-Other White-Hispanic-Hispanic-Asian-Asian-African-American-African-American

33 (38%)33 (38%)48 (55%)48 (55%)4 (5%)4 (5%)2 (2%)2 (2%)1 (1%)1 (1%)(N=88)(N=88)

13 (33%)13 (33%)25 (63%)25 (63%)0 (0%)0 (0%)1 (3%)1 (3%)1 (3%)1 (3%)(N=40)(N=40)

20 (42%)20 (42%)23 (48%)23 (48%)4 (8%)4 (8%)1 (2%)1 (2%)0 (0%)0 (0%)(N=48)(N=48)

0.2080.208

Body mass index, Body mass index, kg/mkg/m22

26.9 (N=88)26.9 (N=88) 24.3 24.3 (N=40)(N=40)

31.1 31.1 (N=48)(N=48)

0.0090.009

Systolic blood Systolic blood pressure, mm Hgpressure, mm Hg

111 (N=88)111 (N=88) 110 (N=40)110 (N=40) 115 (N=48)115 (N=48) 0.1080.108

Diastolic blood Diastolic blood pressure, mm Hgpressure, mm Hg

68 (N=88)68 (N=88) 67 (N=40)67 (N=40) 70 (N=48)70 (N=48) 0.1230.123

All data are medians or counts with proportions of the column using fasting laboratory specimens unless otherwise noted. P-values reflect Mann-Whitney U-tests or chi-square tests, as appropriate. Percentages may not total to 100% due to rounding.

RESULTS: Clinical RESULTS: Clinical CharacteristicsCharacteristics

VariableVariable All SubjectsAll Subjects Steatosis Steatosis AbsentAbsent

Steatosis Steatosis PresentPresent

P-P-valuevalue

Fasting glucose, mg/dLFasting glucose, mg/dL 86 (N=79)86 (N=79) 86 (N=36)86 (N=36) 86 (N=43)86 (N=43) 0.1500.150

Fasting insulin, mIU/LFasting insulin, mIU/L 9.7 (N=79)9.7 (N=79) 7.0 (N=36)7.0 (N=36) 12.6 (N=43)12.6 (N=43) 0.0590.059

HOMA-IR, mmol mIU/LHOMA-IR, mmol mIU/L22 2.04 (N=79)2.04 (N=79) 1.50 (N=36)1.50 (N=36) 3.53 (N=43)3.53 (N=43) 0.0330.033

OGTT, 2hr glucose, mg/dLOGTT, 2hr glucose, mg/dL 99 (N=65)99 (N=65) 93 (N=27)93 (N=27) 101 (N=38)101 (N=38) 0.3150.315

OGTT, 2hr insulin, mIU/LOGTT, 2hr insulin, mIU/L 46.3 (N=62)46.3 (N=62) 42.0 (N=25)42.0 (N=25) 53.9 (N=37)53.9 (N=37) 0.2480.248

Presence of IFG, IGT, or Presence of IFG, IGT, or DMDM

16 (24%) 16 (24%) (N=67)(N=67)

2 (7.7%) 2 (7.7%) (N=26)(N=26)

14 (34.1%) 14 (34.1%) (N=41)(N=41)

0.0130.013

Total cholesterol, mg/dLTotal cholesterol, mg/dL 193 (N=88)193 (N=88) 195 (N=40)195 (N=40) 187 (N=48)187 (N=48) 0.4430.443

HDL cholesterol, mg/dLHDL cholesterol, mg/dL 60 (N=87)60 (N=87) 64 (N=40)64 (N=40) 54 (N=47)54 (N=47) 0.0030.003

LDL cholesterol, mg/dLLDL cholesterol, mg/dL 105 (N=85)105 (N=85) 105 (N=40)105 (N=40) 105 (N=45)105 (N=45) 0.7920.792

Triglycerides, mg/dLTriglycerides, mg/dL 97 (N=88)97 (N=88) 92 (N=40)92 (N=40) 107 (N=48)107 (N=48) 0.1550.155

ALT, U/LALT, U/L 17 (N=88)17 (N=88) 15 (N=40)15 (N=40) 20 (N=48)20 (N=48) 0.0030.003

GGT, U/LGGT, U/L 17 (N=81)17 (N=81) 14 (N=35)14 (N=35) 19 (N=46)19 (N=46) 0.0230.023

Elevated ALT or GGTElevated ALT or GGT 9 (10.2%) 9 (10.2%) (N=88)(N=88)

2 (5.0%) 2 (5.0%) (N=40)(N=40)

7 (14.6%) 7 (14.6%) (N=48)(N=48)

0.1390.139

All data are medians or counts with proportions of the column using fasting laboratory specimens unless otherwise noted. P-values reflect Mann-Whitney U-tests or chi-square tests, as appropriate. Percentages may not total to 100% due to rounding.

RESULTS: RESULTS: Laboratory CharacteristicsLaboratory Characteristics

Prevalence of Elevated Prevalence of Elevated ALT or GGT in PCOS ALT or GGT in PCOS

Study GroupStudy Group

0%

5%

10%

15%

20%

25%

30%

35%

Absent Mild Moderate Severe Present

5%

15%

10%

0%

31%*

Grade of Steatosis

*p<0.01 vs. absent

0%

5%

10%

15%

20%

25%

30%

35%

Lean Overweight Obese Overall

3%

10%8%

19%*

BMI Category*p<0.05 vs. lean

p<0.05 vs. moderate

Schwimmer JB et al (2005)

Prevalence of Elevated Prevalence of Elevated ALT in PCOSALT in PCOS

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Perc

ent

N=70

Obese

Overweight

Lean

Subjects by BMI Group

0%

5%

10%

15%

20%

25%

30%

35%

Perc

ent

30%

Elevated ALT

63% Hispanic, 17% White, 10% Black, 10% Asian

US population estimate

Adapted from Schwimmer J, et al. Fertility and Sterility 2005; 83(2):494.

Prevalence of Unexplained Elevation in ALT or Prevalence of Unexplained Elevation in ALT or ASTAST

in the US (NHANES III) (no detectable liver in the US (NHANES III) (no detectable liver disease or excessive alcohol consumptiondisease or excessive alcohol consumption))

Adapted from Clarke JM, Brancati FL, Diehl AM, American Journal of Gastroenterology 2003; 98(5):960-967

0%

5%

10%

15%

Men Women

Perc

ent

5.7%4.6%

0%

5%

10%

15%

Non-HispanicWhite

Non-HispanicBlack

Mexican-AmericanPe

rcen

t

4.8% 4.2%

9.9%

PCOS & persistently elevated PCOS & persistently elevated LFTLFT’’S may already have NASHS may already have NASH

Persistently elevated aspartate (AST) and/or alanine Persistently elevated aspartate (AST) and/or alanine aminotransferase levels (ALT) U/L > 60U/L suggest aminotransferase levels (ALT) U/L > 60U/L suggest the presence of nonalcoholic steatohepatitis (NASH)the presence of nonalcoholic steatohepatitis (NASH)

The above were present in 29/200 young women with The above were present in 29/200 young women with PCOS (15%) in a Duke University study (Dr. Ann PCOS (15%) in a Duke University study (Dr. Ann BrownBrown’’s Group)s Group)

Liver Bx in 6 such women had Bx-demonstrated Liver Bx in 6 such women had Bx-demonstrated evidence of NASH with fibrosis. MS signs were higher evidence of NASH with fibrosis. MS signs were higher in this group than those without liver Bx and nl LFTin this group than those without liver Bx and nl LFT’’s. s. Mean levels of AST and ALT were 143 and 144 U/L, Mean levels of AST and ALT were 143 and 144 U/L, respectively.respectively.

Setji TL et al. J Clin Endocrinol Metab 2006;91:1741Setji TL et al. J Clin Endocrinol Metab 2006;91:1741

VariableVariable All All SubjectsSubjects

Steatosis Steatosis AbsentAbsent

Steatosis Steatosis PresentPresent

P-P-valuevalue

Presence of Presence of splenomegalysplenomegaly

6 (6.8%) 6 (6.8%) (N=88)(N=88)

1 (2.5%) 1 (2.5%) (N=40)(N=40)

5 (10.4%) 5 (10.4%) (N=48)(N=48)

0.1420.142

Presence of Presence of hepatomegalyhepatomegaly

4 (4.5%) 4 (4.5%) (N=88)(N=88)

0 (0%) 0 (0%) (N=40)(N=40)

4 (8.3%) 4 (8.3%) (N=48)(N=48)

0.0620.062

Presence of PCO Presence of PCO morphologymorphology

64 (72.7%) 64 (72.7%) (N=87)(N=87)

28 (70.0%) 28 (70.0%) (N=40)(N=40)

36 (76.6%) 36 (76.6%) (N=47)(N=47)

0.4870.487

All data are medians or counts with proportions of the column using fasting laboratory specimens unless otherwise noted. P-values reflect Mann-Whitney U-tests or chi-square tests, as appropriate. Percentages may not total to 100% due to rounding.

RESULTS: ImagingRESULTS: Imaging

Steatosis in Non-obese PCOS Subjects Compared to Steatosis in Non-obese PCOS Subjects Compared to ControlsControls

US study of NAFLD Prevalence in PCOS US study of NAFLD Prevalence in PCOS (Cerda (Cerda ’’07)07)

A recent study of 41 Chilean women with PCOS A recent study of 41 Chilean women with PCOS (mean age 24.6 and BMI 30.3) noted an US (mean age 24.6 and BMI 30.3) noted an US prevalence of NAFLD in 41.5% compared to 19% of prevalence of NAFLD in 41.5% compared to 19% of 31 controls31 controls

9 of 17 with NAFLD (64%) had ALT levels >25 9 of 17 with NAFLD (64%) had ALT levels >25 (mean 27.8 IU/L) vs 15.5 in controls ( ?? Increased (mean 27.8 IU/L) vs 15.5 in controls ( ?? Increased NASH)NASH)

17 PCOS women with steatosis had mean ALT of 17 PCOS women with steatosis had mean ALT of 38.9 vs 20.0 IU/L in the 24 nonsteatotic PCOS 38.9 vs 20.0 IU/L in the 24 nonsteatotic PCOS subjectssubjects

IR was noted in 63% (Insulin, HOMA index) vs IR was noted in 63% (Insulin, HOMA index) vs 35.5% of nonsteatotic PCOS35.5% of nonsteatotic PCOS

Similarly, increased WHR was increased in women Similarly, increased WHR was increased in women with NAFLDwith NAFLD

Cerda C et al, J Hepatology 2007;47:412 Cerda C et al, J Hepatology 2007;47:412

CONCLUSIONSCONCLUSIONS NAFLD is common in PCOS regardless of BMI

Obesity and insulin resistance are important associated factors (as has been shown in other population studies)

Despite established hepatic steatosis, elevated liver chemistries are relatively uncommon in PCOS

As non-alcoholic steatohepatitis may be a risk factor for the development of cirrhosis, hepatic failure and infrequent hepatocellular carcinoma, the high prevalence of hepatic steatosis in young obese and non-obese women with PCOS and frequent normal liver chemistries is of concern

Further data are needed to establish the role of lifestyle modification, metformin,TZD’s, & others in its treatment

CollaboratorsCollaborators

Sanjiv V KinkhabwalaSanjiv V Kinkhabwala Maya GambarinMaya Gambarin Thomas D SchianoThomas D Schiano Carol A BodianCarol A Bodian Hsu-Chong YehHsu-Chong Yeh

Comparison of HOMA-IR values Comparison of HOMA-IR values between non-obese PCOS Subjects between non-obese PCOS Subjects

and Controlsand ControlsPCOS Women PCOS Women (nonobese)(nonobese)

N= 47N= 47

Normal Controls Normal Controls (nonobese)(nonobese)

N=18N=18

Median HOMA-IR= Median HOMA-IR= 1.401.40

Min, Max = Min, Max = 0.35, 9.500.35, 9.50

= 1.30= 1.30

= 0.62, 2.30= 0.62, 2.30

No statistical comparison of No statistical comparison of HOMA-IR was possible in HOMA-IR was possible in these 2 groups due to too these 2 groups due to too few controls.few controls.

The HOMA-IR values for 3 The HOMA-IR values for 3 steatosis controls were a Median steatosis controls were a Median value of 1.43 with avalue of 1.43 with a

range of 1.14-2.06. The 15 range of 1.14-2.06. The 15 controls without steatosis had a controls without steatosis had a Median and Range of 1.17 and Median and Range of 1.17 and 0.62-2.27. (No HOMA-IR data in 4 0.62-2.27. (No HOMA-IR data in 4 of the 22 control subjects).of the 22 control subjects).

Should NAFLD be Included Should NAFLD be Included in the Definition of in the Definition of

Metabolic Syndrome?Metabolic Syndrome? HOMA-IR >2, oxidative stress, and circulating HOMA-IR >2, oxidative stress, and circulating

adipokines were correlated to ATP III criteria, and adipokines were correlated to ATP III criteria, and to NAFLD in 197 healthy nonobese nondiabeticsto NAFLD in 197 healthy nonobese nondiabetics

NAFLD more accurately predicted IR then ATP NAFLD more accurately predicted IR then ATP III criteria (including studies of circulating III criteria (including studies of circulating adipokines)adipokines)

Adding NAFLD to ATP criteria significantly Adding NAFLD to ATP criteria significantly improved diagnostic accuracy for IR, independent improved diagnostic accuracy for IR, independent of adiposity, or any feature of the MS in IR of adiposity, or any feature of the MS in IR subjects subjects

Musso G et al , Diabetes Care Musso G et al , Diabetes Care

2008;31:5622008;31:562

Future Investigations for Rx Future Investigations for Rx NAFLDNAFLD

Insulin sensitizersInsulin sensitizers

MetforminMetformin

TZDTZD’’s- piaglitazones- piaglitazone

Cochrane Database System Cochrane Database System Review Review

2007 (1-24-07), Angelico F et al 2007 (1-24-07), Angelico F et al (Rome)(Rome)

0

10

20

30

40

50

60

70

80

90

1st Qtr 2nd Qtr 3rd Qtr 4th Qtr

EastWestNorth

Prevalence and Grade of Steatosis in Prevalence and Grade of Steatosis in Overweight and Lean PCOS Women Compared Overweight and Lean PCOS Women Compared

to Control Subjectsto Control Subjects

Age and the MS:Age and the MS:PCOS and Women of NNANES IIIPCOS and Women of NNANES III

Prevalence of Steatosis by BMI of PCOS Prevalence of Steatosis by BMI of PCOS WomenWomen

RESULTSRESULTS: Prevalence of : Prevalence of Steatosis by BMI Group Steatosis by BMI Group

(off Rx(off Rx))

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Lean Overweight Obese Overall

Per

cen

t

Mild Moderate Severe

53%

75%

63% 59%

N=17 N=4 N=11 N=32

Higher BMI group was associated with increasing grade of steatosis, stratifying on treatment with OCP, metformin, neither, or both (Mantel-Haenzsel test, p=0.0005).

MS and NAFLD in 4400 Japanese MS and NAFLD in 4400 Japanese PatientsPatients

A longitudinal study of 4,401 Japanese men and women A longitudinal study of 4,401 Japanese men and women (aged 21-80 yrs, mean age 47 yrs, BMI 23-25 kg/m2), a (aged 21-80 yrs, mean age 47 yrs, BMI 23-25 kg/m2), a causal relationship between the MS and NFLD was notedcausal relationship between the MS and NFLD was noted

US evidence of NAFLD was found initially in 18% of US evidence of NAFLD was found initially in 18% of participants. After 15 months re-exam participants. After 15 months re-exam ->->additional 10% additional 10% was noted where no NAFLD was present previously.was noted where no NAFLD was present previously.

1/3 met the ATP III criteria for the MS, and observed that 1/3 met the ATP III criteria for the MS, and observed that MS had a 4-11 fold higher risk for NAFLD. In this ethnic MS had a 4-11 fold higher risk for NAFLD. In this ethnic group a small increase in weight of 2-3 Kg exerted group a small increase in weight of 2-3 Kg exerted significant metabolic effectssignificant metabolic effects

They also demonstrated that minimal weight loss and They also demonstrated that minimal weight loss and lifestyle modification was associated with a regression lifestyle modification was associated with a regression NAFLD in almost all with no prior MS. Conclusion: NAFLD NAFLD in almost all with no prior MS. Conclusion: NAFLD is hepatic expression of MSis hepatic expression of MS

Hamaguchi M et al, Ann Int Med Hamaguchi M et al, Ann Int Med 2005;143:7222005;143:722