2004 Summer Meeting - Amsterdam - The Pathological Society of ...

121
The correlation between phenotypic expression and tumor
development in colorectal epithelia tumors
T Yao , K Nishiyama , M Tsuneyoshi
Kyushu University, Fukuoka, Japan
The purpose of this study is to clarify the correlation between the phenotypic
expression and tumor development. Colorectal adenomas and adenocarcinomas
were randomly selected, including 236 tubular (15 low-grade adenoma, highgrade
adenoma 34, invasive carcinoma 187) and 60 villous (low-grade adenoma
15, high-grade adenoma 23, invasive carcinoma 22) tumors. By the
combination of expression CD10, MUC2 and human gastric mucin, the lesions
were classified into five types, large intestinal (LI), mixed (Mix), small
intestinal (SI-type), gastric (G-type) and unclassified (UC) types.
Villous tumors revealed higher incidence of Mix-type (60% in low-grade
and 83% in high-grade adenomas) than tubular tumor did (13% in low-grade
and 6% in high-grade adenomas). On the other hand, tubular tumors contained
SI-type (35% in high-grade adenoma and 22% in invasive carcinoma),
however, no villous tumors revealed SI-type. In tubular tumors, polypoid
growth carcinomas revealed similar phenotypic expression to tubular
adenomas, but they revealed different phenotypes from non-polypoid ones. In
addition, higher incidence of venous invasion was seen in SI-type.
In conclusion, the different carcinogenetic pathway is suggested between
villous and tubular tumors and between polypiod and non-polypiod carcinomas.
Such phenotypic classification is also useful for evaluation of the biological
behavior.
122
Audit Of Compliance With Upper GI Cancer Minimum
Datasets
A BISWAS , KP WEST
LEICESTER ROYAL INFIRMARY, LEICESTER, United Kingdom
Minimum datasets assist pathologists in providing clinically useful and relevant
information in surgical pathology reports. The purpose of this study was to
audit the compliance of upper GI cancer reports with the minimum datasets for
oesophageal and gastric cancers.
We included all upper GI cancer resection specimens reported during the period
2001 to 2003 at Leicester. There were 225 cases, comprising gastric
adenocarcinomas (56%), oesophageal adenocarcinomas (20.2%), gastro
oesophageal adenocarcinomas (14.8%), oesophageal squamous cell carcinomas
(7%) and others (2%). In spite of the free text style of reporting, pathologists
were generally compliant with the relevant minimum datasets. We assessed the
impact of using minimum datasets by comparing the present group with a
randomised collection of 100 cases from 1998.
Overall, present day surgical pathology reports are more likely to contain
prognostically relevant data like circumferential margin status in lower
oesophageal cancers (98% versus 82%), lymphovascular invasion (84% versus
72%) and character of invasive margin in gastric cancers (70% versus 53%). A
significant improvement in the prognostic value of the reports was readily
apparent with the incorporation of minimum datasets in reporting of upper GI
cancers.
123
Dysplasia Risk and Subtypes of Intestinal Metaplasia in
Barrett’s Mucosa
LJ Neilson , RC Stuart , JJ Going
University of Glasgow, Glasgow, United Kingdom
Dysplasia is strongly associated with intestinal metaplasia (IM) but not with
non-intestinal metaplastic mucosal phenotypes (cardiac, fundic) which also
occur in Barrett’s mucosa. This study sought to investigate whether the risk of
dysplasia in Barrett’s metaplasia varies between IM subtypes (type 1 or
complete IM and the two subtypes of incomplete IM defined by the absence
(IIa) or presence (IIb) of sulphomucins in columnar cells).
Biopsies from a total of 812 oesophageal sites (excluding the OG junction) in
long-segment Barrett’s oesophagus patients were stained with High Iron
Diamine / Alcian Blue pH 2.5 to reveal acidic and sulphated mucins and were
scored for the presence of type I IM (with a brush border and negligible
alcianophilia of columnar cells), type IIa IM and type IIb IM (as defined
above). Dysplasia was grouped as either absent (no atypia or reactive only)
versus probable and definite dysplasia. The table shows observed and (exected)
numbers of biopsies by dysplasia category and IM type:
IM type No dysplasia Dysplasia present
I 74(60) 2 (16)
IIa 222 (214) 48 (55)
IIb 376 (397) 123 (102)
2 = 21.4 P90% of the
patients with AC are still diagnosed at first endoscopy.
We reviewed coded diagnoses of all oesophageal biopsies between 1996 and
2003.Surveillance patients are defined as BO or dysplasia on first biopsy with
at least one further biopsy within 3 years. Index AC are diagnosed at first
biopsy, or early repeat biopsy. Surveillance AC are detected in a biopsy during
surveillance.
Results: 1674/5624(30%) biopsies showed BO. 300/980(30.6%) patients with
BO had surveillance, averaging 3.3 endoscopies each. Histology in first biopsy
and worst histology in subsequent biopsies are shown in the table.
First histology Worst histology in subsequent biopsies
BO LGD HGD AC
BO (n=265) 228 25 6 6
LGD (n=28) 11 15 1 1
HGD (n=7) 0 1 6 0
Only 7 of 217(3.3%) patients diagnosed with AC were under
surveillance.6/7(86%)surveillance cases were stage 1 compared with
12/65(18.5%) of the resesctable index AC. Only 2/7 patients with surveillance
AC had previous dysplasia;5 AC were diagnosed after1-3 previous biopsies
showing BO only.
Conclusion: Only 3.3% AC were diagnosed during BO surveillance; we have so
far seen a BO-dysplasia-AC sequence in only 1 patient in 8 years despite
increased frequency of endoscopy in patients with dysplasia.
57