5/26/2017

1

Early Neoplasms of the Upper GI Tract: Classification, Clinical Significance and Management

Gregory Y. Lauwers, MD Senior Member

H. Lee Moffitt Cancer Center & Research Institute Tampa, FL

Gregory.Lauwers@Moffitt.org

“I have no relevant relationships requiring disclosure”

[Inflammation] Dysplasia Advanced CAEarly CA

Riddell R. 1983; WHO, 2010

� Dysplasia (IEN) is defined as an unequivocal neoplastic lesion showing structural and cytological abnormalities confined to the basement membrane. It results from genetic clonal alterations and is predisposed to progression.

� Early carcinomas are limited to the mucosa, or at least, invade minimally the submucosa, lesions without potential (or limited) risk of developing lymph node metastases.

Inter-observer agreement in assessing dysplasia

� Overall: fair (ĸ ~ 0.3)� Indefinite & LGD: poor to fair (ĸ ~0.0-0.3)� Negative & HGD: good (ĸ~ 0.3-0.5)� Similar agreement for GI specialists & generalists� Weaknesses in study design

� Lack of “real life” histological context� Assessment of precision, not accuracy (outcomes)

� Technical (fixation, processing, sampling)

� Overlapping features of dysplastic and reactive changes (active inflammation, post-inflammatory regeneration)

� Synthesis of several histologic parameters which may have conflicting significance

� Histological diversity of dysplastic changes

Impediments to accurate grading of dysplasia

5/26/2017

2

Early Neoplasms of the Upper GI Tract

• Squamous intraepithelial neoplasia

• BE: classic and phenotypic variants of dysplasia

• Gastric dysplasia and early gastric cancer

Esophageal carcinoma

Age-standardized death rates from esophagus cancer by country (per 100,000 inhabitants), WHO, 2010

5

4

3

2

1

0

ADENOCARCINOMA OF THE DISTAL ESOPHAGUS

[US]

lamina propria lamina propria lamina propria lamina proprialamina propria

NormalEsoph.

Regenerative epithwith atypia

LG- IEN HG IEN/CIS

InvasiveCarcinoma

• Progression: 25% & 75% of pts with LGIEN and HGIEN develop inv. CA within10 yrs (5% for patients w/ esophagitis)

LGIEN HGIEN Invasive canormal CIS

Esophageal intraepithelial neoplasia

Structural features Cytological features

Increased cellular density Nuclear enlargement and pleomorphism

Intercellular edema Nuclear overlaping

Loss of differentiation/ surface maturation

Dyskeratosis

Loss of cellular polarity Hyperchromasia

Regular and irregular neoplastic buds Conspicuous nucleoli

Sharp demarcation /Oblique line Increased and atypical mitoses

Kobayashi M, et al. Oncology 2006; 71: 237-245

5/26/2017

3

Diagnostic clue : oblique lineNeoplasticNon-neoplastic

Oblique line, increased cellular density, loss of polarity

Loss of polarity & nuclear atypialamina propria

LGIEN

5/26/2017

4

lamina propria

HGIENlamina propria

CIS

Carcinoma in situ Basal layer type Ki-67

p53

Histology IHC+ Mutation+ Mutation+/IHC+

Mutation-/IHC+

HGD/CIS 86% 71% 83% 17%

LGD 81% 67% 82% 18%

RAE 0% 0% 0% 0%

NE 0% 0% 0% 0%

Proportions of (+) p53 IHC and p53 mutation show no differences between HGD/CIS & LGD

p53 IHC & p53 mutation of esophageallesions

5/26/2017

5

lamina propriaInvasive CA. Superficially invasive Superficially invasive

Makutuchi H, et al. Rinsho Shoukakinaika12: 1749-1756, 1997.

Lymphatic invasion & LN metastasis in esophageal carcinoma

M1 M2 M3 SM1 SM2 SM3Lymphatic permeation 0% 5% 35% 54% 72% 91%

Lymph node metastasis 0% 0% 8% 17% 28% 49%

EPLPMMMSM

MP

T1a-EP(M1)

T1a-LPM(M2)

T1a-MM(M3)

SM1(SM1)

SM2(SM2)

SM3(SM3)

MM

LPM

EP

SM

T1a-LMP (M2)

5/26/2017

6

MM

T1a-MM (M3) Other indication for surgery:�Lymphatic permeation�Positive deep [vertical] margin

Low grade IEN or reactive atypia?

• Enlarged nuclei• Little nuclear variety in

shape • Intraepithelial lymphocytes• Limited within the lower

half• No oblique line

Re-biopsy after 4 w of PPI

Basal cell Hyperplasia: proliferation > 15 % of the total thickness of epith.

• High cell density, but N polarity is maintained and atypia is absent.

• Reactive change reflecting disturbances in maturation of squamous epithelium.

Pseudoepitheliomatous hyperplasia

5/26/2017

7

Management of early squamous neoplasmsSize & endoscopic findings Histology Management

< 5 mm or regular unstained area

Non-neoplastic LG-IEN

Follow up

5 - 10 mm, regular unstained area

Mostly LG-IEN Endoscopic resectionF-up q. 6 months Biopsy, at least in number

5 – 10 mm,irregular unstained area

LG-IENor HG-IEN/CIS

Endoscopic resectionFollow up

> 10 mm,suspicious of carcinoma

HG-IEN/CIS Endoscopic resection

Any size or shape Invasive SCC Surgical resectionChemo / radiation

• Resemble sporadic adenomas

• Lack of maturation & top-down pattern

• Limited glandular distortion

• Pencil-shape hyperchromatic nuclei

• Limited nuclear stratification

• Reduced differentiation

• Increased proliferation

• Clonal character

• Range of epithelia from truly normal, normalized, reactive and regenerative

• Diagnosis of dysplasia is based on recognition of what is perceived as not dysplasia

• Marked architectural distortion

• Lack of maturation • Marked nuclear

stratification• Marked nuclear

enlargement• Marked nuclear

hyperchromasia• Mitoses onsurface• Reduced differentiation• Increased proliferation• p53 overexpression or

loss

HIGH-GRADE DYSPLASIALOW-GRADE DYSPLASIANEGATIVE

Unequivocally neoplastic lesions

indefinite low gradeno high grade

Normal Barrett Esoph. LGD HG D IntramucosalCA

SQUAMOUS EPITHELIUM INTESTINALIZED EPITHELIUM

5/26/2017

8

Corley DA . Gastroenterology ‘02

Age at dx of BE: 61.4yrs

Age at cancer dx:66.4yrs

Age at dx of BE:72.6yrs

Age at cancer dx:78.9yrs

Importance of early detection Interobserver variation

4 LGD18 HGD2 Intramucosal CA

My diagnosis: High grade dysplasia

Montgomery E. Hum Pathol 2001

Interobserver variation3 TINED3 Indefinite8 LGD9 HGD1 Intramucosal CA

My diagnosis: intramucosal CA

Montgomery E. Hum Pathol 2001

5/26/2017

9

High-grade dysplasia

Why is this high-grade dysplasia?

Histologic feature CA in resection

Cribiform/solid growth 73% (33/45)

Dilated tubules/necrotic debris

79% (23/29)

Ulcerated HGD 83% (19/23)

Neutrophils in dysplasia 80% (16/20)

Invasion of sq. epith 100% (5/5)

None of the above 0 (0/16)

1 of the above 39% (7/18)

2 of the above 83% (10/12)

3 of the above 87% (13/15)

4 of the above 88% (7/8)

Zhu W. Am J Clin Pathol 2009;132

Retrospective 127 esophago-gastrectomies performed for HGD, or HGD suspicious for CA [CA present in17% of HGD and 74% of HGD/S cases]

Low-grade dysplasia

5/26/2017

10

Reactive changes or dysplasia? Reactive changes mimicking dysplasia

Stromal inflammation / maturing granulation tissue

Scattered intra-epithelial inflammation

Nuclear and cellular uniformity

Surface maturation

HGD in the bed of ulcerated BE

Various types of dysplasia

non-Adenomatous Type 2 [foveolar]Serrated

5/26/2017

11

Prevalence:6.7% (goblet cells in 62%)94% assoc. typical dyspl. [HGD:+]High rate DNA abnormalities

Non-adenomatous type dysplasia

Maximum dx upon Follow-up

Dysplastic Variant

N Low-grade High-grade Carcinoma

Nonadenomatous 18 0% 78% 17%

Adenomatous 24 25% 54% 21%

Low-grade 13 46% 31% 23%

High-grade 11 82% 18%

Rucker-Schmidt et al. Am J Surg Pathol2009;33(6):886-93

10 Year Follow-up

Non-adenomatous type dysplasia in BE

Modern Pathology; 2010-23:834-843

Prevalence:46%.(HGD:58%) (adjacent IM: 53%)

Adenomatous & Hybrid Dysplasia: Prevalence:27%.(HGD:91%;100%) (adjacent IM: 100%/82%%)HGDLGD

(41 resections w/ dysplasia w or w/o associated inv. ACA)

Progression to cancer of various types of dysplasia

Dysplasia

Association with

Progression to cancer

ConventionalLGD

ConventionalHGD

Conventional LGD (N=22)

1 (5%)

Conventional HGD (N=16)

12 (75%)

Foveolar Dysplasia (N=17)

4(24%) 13(76%) 8 (47%)

Serrated Dysplasia (N=6)

3(50%) 3(50%) 3 (50%)

Srivastava et al, USCAP 2010

5/26/2017

12

Basal Crypt DysplasiaBasal Crypt Dysplasia

• Prevalence:7.3%• 87% have prior or concurrent dysplasia or CA• Association particularly significant w/ regard to the assoc. w/ HGD (P=0.004).

Can we-reliably-recognize BCD?• 40 bx: 10 BE,9 BCD,10 LGD,9

HGD,2 IMCa [selected by the index

pathologist]– 5 (blinded) GI pathologists. – K for IOV for entire cohort

:0.44 (moderate) • [IMC (K=0.65)-LGD (K=0.31)]

Metaplastic atypiaMetaplastic atypia

• No differences in reproducibility of Basal Crypt Dysplasia (K=0.44)-LGD (K=0.31) or HGD (K=0.46)

• When disagreement w/ index diagnosis regarding assessment of BCD (n=17/45 readings), most diagnosed either LGD or HGD rather than BE w/o dysplasia.

Coco et al, 2011 Am J SurgPathol

DNA abnormalities in basal crypt cells

Compared w/ BE, BCD shows:↑ prevalence rate of p53 positivity

(60% vs.13%, P<0.02)↑ total & basal crypt Ki-67

proliferation rate (P<0.001) (similar to LGD or HGD)

Clonal identity (CDKN2A mutations)

Zhang X Am J Surg Pathol ’08; Lomo LC Am J Surg Pathol ’06; Khan S. J. Pathol 2013;231; Srivastava A. USCAP 2011

Molecular anomalies & natural history of basal crypt dysplasia

Recurring issue w/ basal crypt dysplasia

5/26/2017

13

Level 1 Level 2Level

2

Barrett Esophagus Surveillance

• NO dysplasia– Endoscopic surveillance at intervals of 3 to 5 years. – (Repeat at 1 year after 1st first diagnosis not required)

• Indefinite for dysplasia– Repeat endoscopy after optimization of acid suppressive medications for 3–

6 months– If the indefinite for dysplasia reading is confirmed on repeat examination, a

surveillance interval of 12 months is recommended

• Confirmed LGD (& no life-limiting comorbidity)– Endoscopic therapy is considered as the preferred treatment modality– Endoscopic surveillance every 12 months is an acceptable alternative

• Confirmed HGD– Endoscopic therapy unless they have life-limiting comorbidities

ACG Clinical Guideline: Diagnosis and Management of Barrett’s Esophagus 2015

• May arise in:– Atrophic gastritis w/ metaplasia (H. pylori,

autoimmune)– Normal mucosa– Fundic gland polyps (up to 50% in AFP pts)– Hyperplastic polyps (2-16%)– Peutz-Jeghers polyps (2-3%)

Gastric Epithelial Dysplasia

Low grade dysplasia High grade dysplasia

5/26/2017

14

Low Grade Dysplasia

High Grade Dysplasia

Atrophic gastritisIntestinal metaplasiaLow-grade dysplasiaHigh-grade dysplasia

De Vries, Gastroenterology 2008;134

~29%

~3%

5/26/2017

15

Prevalence of Submucosal and LVI in HGD

LGD (n=4) HGD (n=78) HGD+ACA (n= 4) ACA (n=35)

SubmucosalInvasion

0 3 (3.8%) 3 (75%) 4 (11.4%)

Vascular Invasion

0 1 (1.3%) 3 (75%) 1 (2.9%)

Lymphatic Invasion

0 2 (2.6%) 3 (50%) 1 (2.9%)

Sakurai U et al. AJSP 2015

� The low rate of progression indicates that the curr ent grading system does not significantly under-evaluat e the malignant potential of gastric dysplasia, but d oes not fully exclude the risk of submucosal and lymphovascular invasion

Intramucosal CA

breach of basement membraneinvasion into the lamina propria IMC without desmoplasia

“lacy” pattern“cribiform” pattern“disunion” pattern

Intramucosal Cancer Without Desmoplasia• 69-year-old male with dyspepsia

5/26/2017

16

61 62

Your Opinion?

Is this….1.Intestinal Metaplasia? 2.I do not know, but I am not worried?3.I do not know, but I am worried?4.Well differentiated adenocarcinoma?

Is this….1.Intestinal Metaplasia? 2.I do not know, but I am not worried?3.I do not know, but I am worried?4.Well differentiated adenocarcinoma?

Mod Pathol 2013

64anastomosing glands.

branching glandstortuous glands

budding glands

5/26/2017

17

Preoperative bx diagnoses of very well differentiated adenocarcinoma (n=18 patients)

– Adenocarcinoma (36%), – Suspicious for adenocarcinoma (14%) – Indeterminate for neoplasia (21%) – Reactive IM (29%)

Delay in diagnosis :• <1 month in13 (72%);1.5 month (x1);3 mos (x1)(6%). • 3 cases : 23, 50, 84 mos delay w/ minimal changes in size. • Resections revealed that all 3 cases remained IMCs

No endoscopically defined lesion w/ endoscopically defined lesion

LGD Follow- up within 1 year Endoscopic resection to be considered + annual surveillance & bx of any lesion [+/- stop at 2 yrs]

If confirmed: continued surveillance EMR is important in obtaining accurate diagnosis which is upgraded to HGD or CA in up to 19% of cases diagnosed on pinch biopsies

If negative, surveillance (? how long ?) [2 set of negative bx usually conclude but close surveillance]

HGD Immediate endoscopic reassessment with extensive sampling and surveillance at 6-month to 1-year intervals (when to stop?)

Endoscopic resection to confirm Dx & get negative margins + surveillance at 6 moto 1yr intervals w/ bx of any lesion (when to stop?)

Virchows Arch 2012: 460:19–46; Gastroenterology 2008;134

Gastric Dysplasia Surveillance

Thank You!