1

Management of AsymptomaticManagement of Asymptomatic Pancreatic Cysts

Manoop S. Bhutani, MD, FACG, FASGE, FACP, AGAF

Professor, Gastroenterology, Hepatology and NutritionUT MD Anderson Cancer Center, Houston, Texas

Why Asymptomatic pancreatic cysts are important for us?

■ Clinical dilemma

■ High prevalence

1% of patients undergoing abdominal CT scan have a cystic lesion in the pancreas

■ Clinical dilemma

Mucinous vs Non-mucinousBenign vs Premalignant or

Malignant

2

Classification

No epithelium Pseudocyst

Serous epithelium

Degenerative h i lid

y

Mucinous epitheliumMucinous cystic neoplasm

epitheliumSerous cystic neoplasm

Squamous epitheliumL h ith li

change in solid tumorCystic ductal adenocarcinomaCystic neuroendocrine tumorcystic neoplasm

Intraductalpapillary mucinous neoplasm

Lymphoepithelial cyst

tumorCystic acinar cell carcinoma

Serous Cystic Neoplasm (SCN)

F>M

6th decade

Sex

Age 6 decadeAge

Anywhere Location

Multicystic/honeycomb appearance.Central scar and calcification.

Morphology/imaging

Glycogen-rich cuboidal cells lining cystic spaces and amorphous stroma.PAS stain (+)

Histopathology

Generally Benign

Malignancy potentialThirabanjasak D, et al. Pancreatology 2009.

3

Varieties of SCNMicrocystic SCN Oligocystic SCN

F (95%)>>M

4th 5th decade

Sex

Age

Mucinous Cystic Neoplasm (MCN)

4th , 5th decadeAge

Body (tail) Location

Unilocular or septatedcyst +/- wall calcifications.MPD communication (-)

Morphology/imaging

Variable cellular atypiaand secretes mucin with ovarian-like stroma.

Histopathology

Yes Malignancy potentialTanaka M, et al. Pancreatology 2006.

4

Varieties of MCN

M (70%) > FSex

Intraductal Papillary Mucinous Neoplasms (IPMN)

6th and 7th decadeAge

HeadLocation

Dilated main duct or pancreatic ductbranches.

Morphology/imaging branches.Solid component (+) →malignancy?

YesMalignancy potential

5

Intraductal Papillary Mucinous Neoplasms (IPMN)

Mucinous secretion

Fish-mouth appearance

MAIN DUCT IPMN

IPMN

Risk factors for malignancy

■Old■Older age

■Presence of symptoms

■Involvement of main pancreatic duct

■Dilation of main pancreatic duct

■Mural nodule

■Size >30mm (side branch type)

6

IPMN

Differentiation of subDifferentiation of sub--type is important type is important

Side Branch IPMN Main duct IPMN

Prevalence of Cancer Prevalence of Cancer

Low (6~46%) High (57~92%)<<

IPMN

Side Branch IPMN

PD

■ Mucinous cyst■ Communicating with PD

■ Without MPD dilation

7

IPMN

Main duct IPMN

MPD

■ Mucinous cyst■ Dilation of PD (>10mm)

Mucinous cystic neoplasms

Differentiation is important for managementDifferentiation is important for managementCharacreristic MCN Branch duct IPMN

Gender (% female)

Age (decade)

Location (% body/tail)

Common capsule

Calcification

Gross appearance

>95%

4th and 5th

95%

Yes

Rare, curvilinear, in the wall of cyst

Orange-like

~30%

6th and 7th

~30%

No

No

Grape-like

Ovarian-type stroma Positive (76%) Negative

Tanaka M, et al. Pancreatology 2006.

Internal structure

Pancreatic duct communication

Cyst in cyst

Infrequent

p

Cyst by cyst

YesMain pancreatic duct Normal or deviated Normal

8

Imaging diagnosis

■ Computed Tomography (CT)

■ Magnetic Resonance Image (MRI)

■ Endoscopic Ultrasound (EUS)p ( )

Imaging diagnosis -CT-

■ Typical characterization on CT scanS h i DV t l R di l 2006

Unilocular Macrolocular MicrolocularSahani DV, et al. Radiology 2006.

Pseudocyst IPMN, MCN SCA

9

Imaging diagnosis -MRI-

■ Communication with pancreatic duct： IPMN■ Communication with pancreatic duct IPMN

■ Presence of central scar: SCN

MRI may be a feasible tool for surveillance without radiation exposure.

Imaging diagnosis

■ Specific diagnosisp g

CT: Accuracy 61.9 to 76.2%

MRI: Accuracy 55.6% to 76.2%

CT+MRI: Accuracy 73.0% to 77.8% (Lee HJ et al Clinical Radiol 2011)(Lee HJ, et al. Clinical Radiol 2011)

10

Imaging diagnosis -EUS imaging-

HighHigh--resolution images resolution images

Gress F, et al. Am J Gastroenterol 1997.

Koito K, et al. Gastrointest Endosc 2006.

EUS classificationThick wall type Tumor protruding type Thick septal type

Micro cystic type Micro cystic typeThin septa type Simple type

Koito K, et al. Gastrointest Endosc 2006.

11

Imaging diagnosis -EUS imaging-

■ Benign vs Malignant

■ Mucinous vs Non-Mucinous

Accuracy 40% to 93%Accuracy 40% to 93%

Accuracy 51%Accuracy 51%

Thick wallSeptationIntramural nodule/Mass (Ahmad NA, et al. Gastrointest Endosc 2003)

■ Mucinous vs Non Mucinous Accuracy 51% Accuracy 51%

EUS Imaging without FNA not enough as an independent predictor.

Brugge WR, et al. Gastroenterology 2004

Imaging diagnosis -EUS imaging

Cystic fluid analysis with EUSCystic fluid analysis with EUS--FNAFNACystic fluid analysis with EUSCystic fluid analysis with EUS--FNAFNA

■Cytology

■Biochemistry

■Viscosity

■Tumor markers

y

■DNA and others

12

Cystic fluid analysis-Cytology-

■ Mucinous versus non mucinous■ Mucinous versus non-mucinous

■ Mucinous vs Non-Mucinous

Accuracy 59 %

(Thosani N, et al. Dig Dis Sci 2010)

Brugge WR, et al. Gastroenterology 2004.

( , g )

Systemic review and meta-analysis of 11 studiesSensitivity 63 %, Specificity 88%

■ Indirect measurement of mucin, glycoprotein, DNA

Cystic fluid analysis-Viscosity-

contents

■ Viscosity≧1.6 is highly suggestive for mucinous lesion

Brugge WR, et al. Gastroenterology 2004

Linder JD, et al. Gastroenterol Endosc 2006.

Predicted cyst Cyst fluid analysis formula Sensitivity Specificity Accuracy

PC

SCyA

MCyA

MCyA-CA

VIS<1.6 or lipase UL>6000 and CEA ng/ml <480VIS<1.6 and lipase UL<6000

VIS≧1.6 and CEA ng/ml <6000

VIS≧1.6 and CEA ng/ml ≧6000

91.3

100

85.7

100

100

100

100

92

97.2

100

97.2

94.3

13

■ String sign:

Cystic fluid analysis-Viscosity-

■ String sign:

Non-mucinous 0 mm vs Mucinous 3.5 mm (median)Leung KK, et al. Ann Surg Oncol 2009

CPC Study

Cystic fluid analysis-CEA-

C C Study

(Cooperative Pancreatic Cyst Study)

■ Prospectively enrolled 371 patients

■ Histopathology was gold standard

Brugge WR, et al. Gastroenterology 2004.

p gy g

■ EUS, cytology and various fluid tumor markers were evaluated.

14

CPC Study

Cystic fluid analysis--

Receiver operator curve

Sensitivity 73%Specificity 84%

Mucinous vs Non-Mucinous p

analysis of the tumor markers demonstrated that cyst fluid CEA (optimal cutoff of 192 ng/mL) demonstrated the greatest area under the

Accuracy 79%curve (0.79) for differentiating mucinous vs. nonmucinous cystic lesions

Cystic fluid analysis-CEA-

• Pooled analysis of 12 studies, 450 patients.C t ith l t ti 250 U/L• Cysts with an amylase concentration <250 U/L were SCA, MCA, or MCAC (sensitivity 44%, specificity 98%) and, thus, virtually excluded PC.

• A carcinoembryonic antigen (CEA) <5 ng/mL suggested a SCA or PC (sensitivity 50%, specificity 95%)

Van der Waaij LA, et al. Gastrointest Endosc 2005.

95%). • CEA >800 ng/mL strongly suggested MCA or MCAC

(sensitivity 48%, specificity 98%).

15

Cystic fluid analysis-DNA Quantification and Allelic Loss-

PANDA Study

■ 391 patients (Final cohort was 113)

■ Mucinous vs Non-Mucinous

■ Prospective, multicenter trial

■ Malignant vs Pre-malignant

Khalid A, et al. Gastrointest Endosc 2009.

PANDA STUDY

• The study cohort consisted of 113 patients with 40 malignant, 48 premalignant, and 25 benign cysts.premalignant, and 25 benign cysts.

• Cyst fluid k-ras mutation was helpful in the diagnosis of mucinous cysts (odds ratio 20.9, specificity 96%).

• Components of DNA analysis detecting malignant cysts included allelic loss amplitude over 82% (AUC 0.9) and high DNA amount (optical density ratio >10, AUC 0.79). The criteria of a high amplitude k-rasmutation followed by allelic loss showed maximum specificity (96%)mutation followed by allelic loss showed maximum specificity (96%) for malignancy.

• All malignant cysts with negative cytologic evaluation (10/40) could be diagnosed as malignant by using DNA analysis.

16

DNA analysis to the rescue!

Cystic fluid analysis-DNA Quantification and Allelic Loss-

y

1

23

456

Pathological findings

IPMC-NI

MCACLiver met

IPMC-IIPMC-NIIPMC-NI

ALA% if >82%

Yes

Yes

Yes

OD if >10

YesYesYesYes

Cycle threshold value if <25

YesYes

K-ras followed by allelic loss

Yes

Yes

Yes

(Adapted from Ref. 24)

78

9

10

IPMC-NIIPMC-NI

IPMC-NI

IPMC-I

YesYes

Yes

Yes

Yes

Yes

YesYes Yes

The most important clinical tools areThe most important clinical tools are

ACG Guidelines

■ Cross sectional imaging

■ Endoscopic ultrasound

■ Cystic fluid analysis

The most important differential isThe most important differential isThe most important differential is The most important differential is

Mucinous or Non-Mucinous

Khalid A, et al. Am J Gastroenterol 2007.

17

Management

Cost effectiveness for asymptomatic pancreatic cyst Cost effectiveness for asymptomatic pancreatic cyst 26)26)

■ Strategy 2: an aggressive surgical intervention

■ Strategy 1: no specific intervention

■ Strategy 3: EUS-guided FNA with cystic fluid analysis

Compared them in Markov-based clinical model

Das A, et al. Gastrointest Endosc 2009.

Management

Cost effectiveness for asymptomatic pancreatic cyst Cost effectiveness for asymptomatic pancreatic cyst

A strategy based on risk stratification of malignant potential by EUS-guided FNA and cyst fluid analysis is thecyst fluid analysis is the most cost-effective strategy.

Das A, et al. Gastrointest Endosc 2009.

18

Surgery

SCN MCN IPMN

Malignant potential?

No Yes Yes

9 28)

Surgery-MCN-

■ Malignancy: 17.5% 9, 28)

■ The prognosis is excellent

if resected before invasion.

Resection is recommended

Tanaka M, et al. Pancreatology 2006.

Clippa S, et al. Ann Surg 2007.

19

Side-branch IPMN

Surgery -IPMN-

Main duct IPMN

Prevalence of cancer

57% to 92%

Good surgical candidate should be offered surgical Good surgical candidate should be offered surgical resection. resection.

Tanaka M, et al. Pancreatology 2006.

Side-branch IPMN

Surgery -IPMN-

Main duct IPMN

Prevalence of cancer

6% to 46%

Lower prevalence of malignancyLower prevalence of malignancyTreatment should be individualized Treatment should be individualized

Tanaka M, et al. Pancreatology 2006.

20

Surgery -For IPMN-

International consensus guidelineInternational consensus guideline

■ Symptoms attributable to the cyst (e.g., pancreatitis)

■ Dilation of the main pancreatic duct (≥10 mm)

■ Cyst size ≥30 mm

■ Presence of intramural nodules

■ Cyst fluid cytology suspicious/positive for malignancy

Tanaka M, et al. Pancreatology 2006.

Management of asymptomatic cysts

Cyst size

<1<1cm

EUS Monitor

High risk stigmataHigh risk stigmata

MonitorEUS+FNA

Yes No

21

Management of asymptomatic cystsCyst size

1 to 3cm >3cm

EUS+FNA

Consistent with mucinous

Yes

EUS+FNA

Consistent with mucinous

Resection

?

NoHigh risk stigmataPositive cytology

Monitor

No Yes

Yes

Monitor

No

Pancreatic cysts-Injection of alcohol

EUS id d h l l fGan et al, GIE 2005

Dewitt et al, GIE 2009

EUS-guided ethanol lavage lt i t• EUS guided ethanol lavage for

ablation of pancreatic cystic lesions. 23/25 patients with complete follow-up, 8 patients (35%) had complete resolution of their cysts on follow-up imaging

Dimaio et al, Pancreas 2011

results in a greater decrease in pancreatic cyst size compared with saline solution lavage with a similar safety profile.Overall CT-defined

• Two sessions of EUS guided lavage versus only one session.

• 38% had complete resolution after two sessions of ethanol lavage.

Overall CT defined complete pancreatic cyst ablation was 33.3%.

22

Panc.cysts-EUS guided Chemo.

• EUS guided ethanol lavage with Paclitaxel injection in 14 pts cystic pancreatic tumors Oh GIE 2008pts.-cystic pancreatic tumors.Oh,GIE,2008

• Complete resolution in 11/14 patients(78%)

Alcohol injection or alcohol with paclitaxel are promising but not ready for prime time routinepromising but not ready for prime time routine clinical application.

28 months follow up

■ Size <3 cm71 patients

■ Solitary cyst

■ No septationNo solid component

■ Normal PD caliber

45/71 (63%) with f/u studies 26/71 (37%) no f/u studies

31/45 (69%) reviewed 14/45 (31%) not available

■ Normal PD caliber

■ CEA < 192 ng/ml

Pausawasdi N, et al. Surgery 2010.

31/31 (100%) no progression

23

Follow up

Wh t i th ti l i t l f ill ?Wh t i th ti l i t l f ill ?What is the optimal interval of surveillance?What is the optimal interval of surveillance?

2 year follow up from the baseline study for cyst size<3cm, no intracystic growth

Das et al, Am J Gastro 2008

My current practice:First follow up may be at 12 months with MRI and if no change repeat another MRI in one year.

MRI may be better test for follow up than CT

My current practice:

Take home message

■ There are variety of pancreatic cystic lesions.

Mucinous vs Non- mucinous.

Malignant vs Non-malignant.

■ A strategy based on risk stratification of malignant potential by EUS-guided FNA and g p y gcyst fluid analysis is the most cost-effective strategy.

24

The EndThank you for your attentionThank you for your attention