Alpha1-antitrypsin, alpha1-antichymotrypsin, and alpha2-macroglobulin in human gastric carcinomas: A retrospective immunohistochemical study

Alpha -antitrypsin, Alpha -antichymotrypsin, and Alpha2-macroglobulin in Human Gastric Carcinomas: A Retrospective Immunohistochemical Study EIICHI TAHARA,MD, HISAO ITO, MD, KIYOMI TANIYAMA, MD, HIROSHI YOKOZAKI,ME), AND JOTARO HATA,MD attempt to clarify the relation between the expression of these protease inhibitors and prognosis.

One hundred twenty-six gastric carcinomas (68 advanced cancers and 58 early cancers) were examined immunohistochemi. eally for alphat-antitrypsin (AAT), a l p h a l - a n t i c h y m o t r y p s i n (ACT), and alpha2-macroglobulin (AMG) within tumor cells. The incidence of these three protease inhibitors was markedly higher in advanced than in early cancers, regardless of the histologic type of gastric carcinoma. In advanced cancers the incidence of both AAT and AMG was significantly higher in well-differentiated adenocarcinomas than in poorly differentiated adenocarcinomas, but no difference was observed in the expression of ACT between these two types of advanced carcinomas. Eighty per cent of the AAT-positive advanced carcinomas had ACT, and 40 per cent of these tumors also contained AMG. The two-year survival rates clearly indicated that well-differentiated adenocarcinomas with AAT have worse prognoses than well-differentiated adenocarcinomas without AAT, but there was no relation b e t w e e n the e x p r e s s i o n o f ACT or AMG and prognosis. These results strongly suggest that the presence of protease inhibitors in gastric carcinomas is related to the invasive growth of the tumors and that AAT is a tissue tumor marker of welldifferentiated adenocarcinomas of the stomach. It may also serve as a biologic marker of high malignancy in patients with these gastric cancers. Hu.~f PATHOL 15:957--964, 1984.

MATERIALSAND METHODS

It is well known that alphavantitrypsin (AAT), alphal-antichymotrypsin (ACT), and alpha2-macroglobulin (AMG) have protease inhibitory activity. Recently, A A T and A C T were demonstrated by immunohistochemical techniques not only in normal epithelial cells but in tumors of the gastrointestinal tract as wellA -3 However, the effect of the presence of AAT, ACT, or AMG on the biologic behavior of gastric cancers has not been studied in detail. Recently, we examined immunohistochemically the relation between various protein markers of gastric cancers and tile prognoses of the patients. 4-8 Gastric carcinomas producing peptide hormones, !ysozyme, or h u m a n chorionic g o n a d o t r o p i n (HCG) had p o o r prognoses. In the present study the incidence of AAT, ACT, and AMG was investigated in the different stages of various histologic types of gastric carcinomas in an

Received July 19, 1983, from the Department of Pathology, Hiroshima University School of Medicine, Hiroshima, Japan. Revision accepted for publication November 17, 1983. Supported in part by a Grant-in-Aid for Cancer Research from the Ministry of Education, Science and Culture (nos. 57015069 and 58015078). Address correspondence and reprint requests to Dr. Tahara: Department of Pathology, tfiroshi'ma University School of Medicine, Kasumi 1-2-3, Minami-ku, Hiroshima, Japan 734.

One hundred twenty-six gastric carcinomas (68 advanced cancers and 58 early cancers) were studied. All of these gastric tumors were surgically resected during the period from 1975 to 1982 and fixed in 10 per cent b u f f e r e d formalin. For each case two or three sections of the t u m o r were stained with hem a t o x y l i n - e o s i n , periodic a c i d - S c h i f f (PAS) reagent, alcian blue (AB), and high iron diamine (HID). T h e histologic classification of gastric cancers and the definition o f early gastric cancer followed the criteria of the Japanese Research Society for Gastric Cancer. 9 For tile detection of AAT, ACT, and AMG in the tumor tissue, the immtmoperoxidase technique reported previously was employed. 4 Deparaffinized sections (4 tzm thick) p r e p a r e d on albumin-coated slide glasses were treated consecutively for 30 minu t e s with 1) specific rabbit antisera at room temperature, 2) g o a t - r a b b i t IgG antiserum (dilnted 1:50), and 3) soluble peroxidase-antiperoxidase complex (rabbit, 1:50). Sections were stained for five to ten minutes with a solution o f 30 mg of 3, 3'-diaminobenzidine tetrahydrochloride in 100 ml o f 0.05 ,xl TRIS buffer, pH 7.6, containing 0.001 per cent hydrogen peroxide. T h e sections were counterstained with 3 per cent methyl green, AB, and PAS. The AAT, ACT, and AMG antisera purchased from Dakopatts, D e n m a r k were employed at dilutions of 1:150, 1:200, and .1:150, respectively. Alphafetoprotein (AFP) antisera (also purchased from Dakopatts) were employed at a dilution of 1:150. The specificity o f the reaction was d e t e r m i n e d as described by Sternberger. 1~ Surgically resected stomachs from three patients with gastric ulcers and two patients with duodenal ulcers were also examined to determine the localization and distribution of AAT, ACT, and AMG in normal gastrointestinal epithelium. RESULTS Gastric Carcinoma

Tile AAT, ACT, and AMG immunoreactivities of tumor cells from 68 advanced cancers and 58 early 957

PATHOLOGY

HUMAN

T~LE

1.

AAT, ACT, a n d A M G I m m u n o r e a c t i v i t y

Histologic Type*

Stage Advanced (68 cases)

in T u m o r Cells o f 126 G a s t r i c C a n c e r s

No. of Cases with Immunoreactivity

Total No. of Cases

AAT

ACT

AMG

36 27 5 43 15

20 (55.5%) 2 (7.4%) 1 (20.0%) 3 (6.9%) 1 (6.6%)

24 (66.7%) 14 (51.9%) 2 (40.0%) 10 (23.2%) 5 (33.3%)

l 1 (30.6%) 2 (7.4%) 0 2 (4.6%) 1 (6.6%)

Pap, tub Por, sig Muc Pap, tub Por, sig

Early (58 cases)

V o l u m e 15, No. 10 (October 1984)

ABBREVIATIONS:AAT, alphal-antitrypsin; ACT, alphai-antichymotrypsin; AMG, alpha2-macroglobulin. * According to the classification of the Japanese Research Society for Gastric Cancer: pap, papillary adenocarcinoma; tub, tubular adenocarcinoma; muc, mucinous adenocarcinoma; por, poorly differentiated adenocarcinoma; sig, signet ring cell carcinoma.

cancers are summarized in table 1. T h e presence of A A T was observed in 23 advanced cancers (33.8 per cent), but by histologic type A A T was seen in 20 of 36 well-differentiated adenocarcinomas (55.5 per cent; papillary and tubular adenocarcinomas) and in two of 27 poorly differentiated adenocarcinomas (7.4 per cent), including signetring cell carcinomas. T h e frequency o f A A T was significantly greater in advanced, well-differentiated adenocarcinomas than in poorly differentiated adenocarcinomas (P < 0.01). Furthermore, one of the five mucinous adenocarcinomas showed A A T immunoreactivity. On the other hand, in early gastric cancers A A T was observed in three of the 43 well-differentiated adenocarcinomas (6.9 per cent). A statistically significant difference (P < 0.01) was observed in the frequency of A A T in well-differentiated adenocarci-

. . . . . . . . . . .... ,. ~ ~ ~ ...-.~., . ..,

",.7 .

. . . . . . . . . ~. .~ . ~, . . . . .

~

t.

..,,:;

.~

.... ~ - ' ,

".--',

";"

-"

,

-: _ . ; , ~ -

,~:~.,'~,

'

:..,.

~..%,..~,

-

.

9-

-

,:...

.:,:.

-".-,',io,,.v_>

"

..~':

: ",r_-

":

~-~2">

~

-" ~ .

9.'~...

-

~;-~'-

"-

9 -

_-

9

F I G U R E t Cleft), Immunoreactivity.

~

.,:

.-

-J:,~'-',

". , W - :

.

.

..

~

,,

- : , : . = , " :~--,.-, .-:;

~"r.~

-:,~'i'.-"

"

'~ ,~"

Tubular adenocarcinoma [Peroxidase-antiperoxidase

"

.

g. ,: . ~ - ,

...

~"-.

:

-

. ".r"

-

"

;:

~l#,:)"

" -

"

,

-:-

... . ~ z'-

.

.:

."

.-

" . .'~ -

.

:

--,..

l:.l

-" :-"

".f

~"

-4

.,.i-" ,, ~ ,

,,., .

:."

'"

,

-

0~, "

Numerous

from a 49-year-old x 640.)

958

-' '" :--

.--

. - J . - " "

. "-

tumor

man.

--:: __-- :-~--' .

"

{

" .~ll

-g

-.,.'~4,'~. "

. . . .

-

..

-..-

-

-

'

":,

,

"

""

.-

.

~.~

~

.:

""

-

tumor

" /, "

-

-

.'

"-~

".

-"

,.

9

,,.

",.:.#.:y.-.';.,'

;:.-"-/#.... .-

"

:~ : - : ,

o r

. -

f ,'t'.

~

,

..,

"'.-

- -_7'--

- " ,

"

..~'7

,'X,~"

,":,'~

"

;:%

9

9

'~.:

:.."

.

" ~ "

cells show weak ~

Numerous

.

. . . . . . .

. . . .

" .

- ...

~

9

,..'. "" -,

9 . ~ . . - . 4 ,

-:.: " ,: ""

'"r"'/~_ .o

t.

....~'1,.

'-'..''

e .~':;,-_,.~.'5-,.'!I~".." ." lT:'Y "-"-"~ ~ "-~--"-~

: ' 2 -"

,,,',

' : i'

,

,,

.-.

~.'.'~o.."..
9

,,,.-Xr

"~--,,,, ...

~

,~. ~,

~..

" , . ' ~ /

" ~il

."i~ ~~.;

~-.,

.:-" "

%" ,

-

--"S--~;-_

.

.l"~l.

--

:.. t~,-'. ;:, '~_ ,.':..:.:"

from a 7/-year-old woman. technique, x 368.}

Poorly differentiated adenocarcinoma F I G U R E 2 (right). v a r i a b l e intensities. ( P e r o x i d a s e - a n t i p e r o x i d a s e technique,

_

_ .

"--'5--:' W._..'-'~,,

--'.'.'~" -

"..-

.

.c-~":

d;~:ili.''

"

-_ ...-

-r

" : , ~ -:

..~:;~:,-~:..-

"

.

--7

..~ ....

.

.,~.;,.hi-.'-.. <-

-;:.-~.::.~,,:,

""-:~"..-

--

"

.'.:v,:'--,->:, ~,+t 3 ;7

..:~.:-:',:-

..:~

~.=.~..:. "--

. " ;,.'-"

~,

,. ..

-....,..--,.: . . . . . .

(,'~ :~" .-.',-_: ---:. ,..~._-E._',":-.~-" .: -; . . .

'

""

~ ............ -._...- . . . . .

.

.....

:.--...'-.::,'.w.

:?

9

.":,~"~~.-

- .._~.,w.rf.,,.

- L . - - : " ' . . - ~ "-

,~

:

"-~','.'>.,

--~ ............

4"

-~-:. : -

9

-_~:'~:..-~j.

"-

~ "

#.-.-:-. ? ' ~ . - ' - ' j : , .

,~,

9 "

:

,

:,--_:':,~ ..,!~i:-.;-

"~" ~

~ "

"-': ....... L:"I'. .....

!kL <" ~:~-~'i

-

........

" .,..-'." :..~; . : - : . . : --. .-. ~ ' : : ~ - U : ~ ":.'~'-'v" ~ " ~ - : - " - ~ '-,.--"::-":; -"" " ::" "",'4.~'- ~.:-.~'.~"~.".~"." '. ;:.~-W:~

. . - T ~ o ~ -

4 . .

.--:--.,

,.. ; . . . . . .

.. .,.,:,...

- .~'.?;:

-.,:-:~TP_'..:"-

r~:'.{,L-

"-"

_ ~

.............. ~..~ ,, F-x,, , . ~. . ...~. .,

:~,~"

Y.-'-'.:.

--_ -"

:i,~. ':-"-..~"

9

:';..,~s -

-

"-':.'-~

............

"b

9

"

- _.'t'.,,-,,

-

"

~

-

. . . . . .

;.

"~C~-~-:

- . ~ - ~

~ r ;

"

.

c

:

". ~--i,i,~'":,-'~.." -9

,),-.~.,,..

"

_."~".-~:'i.',

"

"/-,-.

9

.

9

~ ' " _ . ; ~ . 9 . ........

.:-.~_~.-,.,g,-~...

nomas between early and advanced cancers. Macroscopically, o f the 23 a d v a n c e d cancers, ten were located in the intermediate zone, eight in the body, four in the antrum, and one in the cardia. T h e mean age of the patients with these cancers was 64.1 years for the women (six cases) and 66.5 years for the men (17 cases). T h e microscopic appearances and numbers of AAT-immunoreactive cells within the tumors varied. Among the 23 advanced gastric cancers, eight welld i f f e r e n t i a t e d a d e n o c a r c i n o m a s (fig. 1) and one poorly differentiated adenocarcinoma (fig. 2) had n u m e r o u s A A T - i m m u n o r e a c t i v e cells. In many o f these cancers AAT-positive cells showed diffuse cytoplasmic staining of variable intensities. In some of the well-differentiated adenocarcinomas, r o u n d or g r a n u l a r cytoplasmic inclusions with "halos" that

"

.

".lli~ "

. . .

,+ z

.

,

9

to strong alpha~-antitrypsin

cells show AAT immunoreactivity

{AAT) of

GASTRIC CARCINOMA [l"ahara et al.]

TABLE 2.

Synchronous Localization of AAT,ACT, and AMG in 20 AAT-positive Well-differentiated Adenocarcinomas (Advanced] Protease Inhibitor hnmunoreactivlty in the Same Tumor

AAT AAT AAT AAT

TABLE 3. Relationship Between AAT Immunoreactivily In Tumor Cells and Inflammatory Reaction in Stroma of 68 Advanced Gastric Carcinomas

Total No. (%)

only + ACT + ACT + AMG + AMG

+ ++ +++

3 (15.0) 8 (40.0) 8 (40.0) 1 (5.0)

showed A A T immunoreact!vity were observed (fig. 3). These cytoplasmic inclusions in the tumor cells showed faint eosinophilia and prominent PAS positivity (fig. 4). However, these inclusions were AFPnegative. For advanced Gincers A C T immunoreactivity was found in 24 well-differentiated adenocarcinomas (66.7 per cent; fig. 5) and in 14 poorly differentiated adenocarcinomas (51.9 per cent). There was no difference in frequency between the two types. In early cancers ACT immunoreactivity was seen in ten welldifferentiated adenocarcinomas (23.3 per cent) and in five poorly differentiated adenocarcinomas (33.3 per cent). Only in well-differentiated adenocarci-

.

+

++

20 10

6 6

2 3

8

2

2

7

+ + +

I

nomas were the frequencie s of ACT and AAT, significantly different (P < 0.01) between early and advanced cancers. The presence of ACT was observed not only in the cytoplasm but also in the nuclei of the tumor cells. For advanced cancers AMG immunoreactivity was found in 11 well-differentiated adenocarcinomas (30.6 per cent) and in two poorly differentiated adenocarcinomas (7.4 per cent). The frequency was also significantly higher (P < 0.01) in well-differentiated adenocarcinomas than in the poorly differentiated tumors. Moreover, as shown in table 1, there was a significant difference (P < 0.01) in the frequency of AMG in well-differentiated adenocarcinomas be-

,'"

J

--

* This reaction was graded _+, + , + + , o r + + + on the basis of the number of neutrophilic granulocytes. t T h i s reaction was also graded -+, + , + + , o r + + + on the basis of the number of positively staining tumor cells rather than the intensity of the staining of individual cells.

ABBREVlATIOXS: AAT, alphan-antitrypsin; ACT, alpharantichymotrypsin; A M G , alpha2-macroglobulin.

".o

A A T Immunoreactivityt

Cell Infiltration Around the T u m o r *

" :'-": "

"~,:",>",

"*

"~'%

4"-'

.

Qq-

'.

> "

'"

~:-'i

"

i

i

*~

|

/L

9

n

. .

-/ 9

,

.

U.

9 .

o.,~

o)

.

~

_ r ~ , ~ -

" "

"7,'C/"%.

= ' ~

-..

) o~

*"~,."

,

*.)

"~- ",

-~,

~t.

~f-~.."...

Same tumor as in figure 1. Round cytoplasmic inclusion with "halo" [Peroxidase-antiperoxidase technique, x 368.]

FIGURE 3 (left).

9

~*:

r.%'..

," ": 9 ,

.._

..~,

lml'l~'q-).

:3".,A)._l~,-/

"

"~"

,.'~ . "

0"

tL~I~.~.

" ,

'.

9 1"4, ~" " .

,~

/~'"

"

I,

(arrow) shows alphat-antitrypsin (AAT] immunoreactivity.

FIGURE 4 [ r i g h t ) . Tumor cells.from .the same area shown in figure 2. Cytoplasmic inclusion containing AAT gives PAS positivity. (Periodic a c i d - S c h i f f stain9 x368.)

959

HUMAN PATHOLOGY -

i

~

;.L"."

'

-;;_:_:~

.

-;..,

.

9

Volume 15, No. 10 [October 1984]

,77/..,_

I

-

,

;

... ~-., ~1

~

.'t.~'~r~ ,~,~,.

,-,'~,.7,~.W%.' .,~C"

,m

,.

"'%

-

_".

.

-

.7

. : ~,:

. .~,,.,

t

"#,

"

/

' . . ,

.-

m

",.

....

_

9 i:'~;;p,

......

~ .

....

.

.-..1-..-

.

.:'...

z,:..-,~k-:::/

.;.~.--.~.:;,~:-

:.-::.'-,

:: , ::-----" --..~',..r : . . . .

-

,.

FIGURE 5 {/eft'}, Same tumor as in figure 1, Many alphal-antichymolwpsin-positive cells are present in another area of the tumor. [Peroxidase-antiperoxidasetechnique9 x 368.) FIGURE 6 (right]. Same tumor as in figure 1. A few tumor s show alpha2-macroglobulin immunoreactivily, (Peroxidase-antiperoxidase technique9 x 368.]

tween early a n d a d v a n c e d cancers. Compared with A A T or ACT-positive cells, AMG-positive cells ap" peared sporadically and were less numerous (fig. 6). In addition, synchronous localization of AAT, ACT, and AMG in advanced gastric cancers was examined. T h e combination of these three markers in

the same t u m o r in 20 AAT-positive well-differentiated adenocarcinomas is summarized in table 2. Sixteen of the AAT-positive tumors (80 per cent) were ACT-positive; eight o f these tumors (40 per cent) were AMG-positive as well. On the contrary, of 14 ACT-positive p.oorly differentiated adenocarci-

%

%

100

100

I

8O

80 . . . . . . .

I I

60

L ......................

40

I

i

60

I

I

; .....

~,

4o

-I

,

j

I 20

20

OP

i

i

i

i

i

I

3

6

9

12

15

18

l

21

U

24

OP

Months

I

l

l

i

I

i

3

6

9

12

15

18

i

21

24

Months

FIGURE 7 (left). Comparison of two-year cumulative survival rates in well-differentiated adenocarcinoma between 18 tumors with and 13 without alphal-antitrypsin [AAT) [advanced]. Solid line, tumors with AAT [mean age, 69.3 years]; dotted line, tumors without AAT [mean age, 70.1 years). FIGURE 8 [right]. (~omparison of two-year cumulative survival rates for well-differentiated adenocarcinomas between 20 tumors with and 11 without alphat-antichymotrypsin (ACT] (advanced]. Solid line, tumors with ACT [mean age, 70.5 years]; dotted line, tumors without ACT [mean age, 68.2 years].

960

GASTRICCARCINOMA [lahara et all % 100

100

80

80

60

60

%

1

"............ 40

40

,,.. . . . . . . . . . . .

L. . . . . . . . . .

i

0P

3

i

i

6

9

1 I

"I,

2O

0

........

i

i

I

20

I

i

i

i

i

I

I

12 15 18 21 24 0P 3 6 9 12_ 15 18 21 24 Months Months FIGURE 9 (left). Comparison of two-year curnulative survival rotes for poorly differentiated oclenocorcinomos between '14 tumors with end '11without ACT. Solid line, tumors with ACT (mean age, 58.4 years); doffed line, tumors without ACT (mean age, 62.'1 years}. FIGURE t0 (fight). Comparison of two-year cumulative survival rates for well-differenfi~ed adenocarcinornas between 2'1 tumors with and fen without oipho2-macroglobulin {AMG]. Solid line, tumors with AMG [mean age, 70.4 yearsl; doffed line, tumors without AMG (mean age, 68.0 years].

from that for patients with AMG-negative well-differentiated adenocarcinomas (47.7 per cent). T h e AMG-positive cases, however, showed the same cumulative survival curve as that of AAT-positive cases (fig. 10) due to the overlap of nine of the 11 cases (81.8 per cent) of AMG-positive tumor with cases of AAT-positive tumor.

nomas, only 11 (78.5 per cent) had ACT. T h e remaining three had ACT, AAT, and/or AMG simultaneously. T h e relation between AAT immunoreactivity and inflammatory reaction in advanced cancers was examined (table 3). However, AAT immunoreactivity was not ahvays correlated with the number of neutrop|fil granulocytes around the tumor cells. We also conducted a follow-up study in 31 of the 36 cases o f advanced cancer and compared the twoyear survival rates for patients with tumors positive and negative for AAT, ACT, and AMG. For 18 AATpositive well-differentiated adenocarcinomas the twoyear survival rate was 33.3 per cent, while that for 13 A A T - n e g a t i v e well-differentiated adenocarcinomas' was 61.5 per cent. The difference in the twoyear survival rates was statistically significant (P < 0.05). T h e r e f o r e , the prognosis for well-differentiated adenocarcinomas w i t h ' A A T was worse than that for well-differentiated adenocarcinomas without AAT (fig. 7). However, as shown in figures 7 and 8, no apparent difference in the two-year survival rate was observed between ACT-positive and ACT-negative tumors, either for well-differentiated or poorly differentiated adenocarcinomas (figs. 8 and 9). In welldifferentiated adenocarcinomas the two-year survival rate was 65 per cent for ACT-positive tumors and 45.5 per cent for ACT-negative tumors. In poorly differentiated adenocarcinomas the rate was 39.8 per cent for ACT-positive and 18.1 per cent for ACTnegative tumors. Prognoses were worse for ACTnegative than for ACT-positive pearl}; differentiated adenocarcinomas. Moreover, the two-year survival rate for patients with AMG-positive well-differentiated adenocarcinomas (40 per cent) was not significantly different

Normal Gastric Mucosa

In tile n o r m a l mucosa of surgically resected stomachs from patients with gastric or duodenal ulcers, AAT immunoreactivity was present in scattered epithelial cells of pyloric and cardiac glands. Moreover, AAT-positive cells were restricted to small portions of these glands (fig. 11). Gastrin-producing G cells and somatostatin-producing D cells did not show any positive reaction against AAT. In the fundic mucosa AAT was detected infrequently in some of.the parietal cells. T h e localization and distribution o f ACT were similar to those o f AAT, but ACT was not present in parietal cells. T h e r e was no AMG immunoreactivity in epithelial cells from the normal gastric mucosa. T h e intestinal metaplasia of the stomach did not contain ACT, AAT, or AMG. Moreover, AAT could not be detected in Paneth cells. On the other hand, numerous AAT-positive epithelial cells were found in the pyloric and cardiac mucosa adjacent to carcinomas (fig. 12). B r u n n e r glands adjacent to the tumors showed strong AAT immun0reactivity (fig. 13). Grading of ACT immunoreactivity was also elevated around tumors. However, AAT and ACT immunoreactivities in epithelial cells around tumors were not related to AAT or ACT expression in the tumors. No AMG was found in the noncancerous epithelium a r o u n d the tumor. Mast cells showed strong AAT immunoreactivity (fig. 11). 961

HUMAN PATHOLOGY

r

,

.,

,_ 9

, , - 0 . . . . , . ~

'L,.''e

"

:

"

",.r

_.

\,

:

.-.,,q

9

.

.

,..~.

5. .... : .:;

, .

-

,-

9

~.":',~,

9

.

....,

9

V o l u m e t5, N o . 10 ( O c t o b e r 1984)

.

.. . . .

Z - '

.

.

~

.

-.-

"'

'

'

,:.,~..

.

, "

.

"

'" "

9

.,.,

~ - , '

,. ",r.-

" ";"* : ' , ~ i ,

.~kf...,.i.., 9".,~._-,.... v ,i.:,

-,

:~- -

'

-.-;~

.

~'- . . . .

9

.

..

;

. "

:

--

".-

.,.,.~..;.-~

-,

.

;.

9

,G',,"

~, Z . . "

t':'(L!

,'~- .;

~ 5",.'.

' ,

~:"k'/ ~"

- '

"~

'-Z'1.

" I;"

".~..

--

~ I.

".,

""'

,

t

'

. /'t-

4

.;:~

" - : - " % ,'~' . - ' , ' -

''-"

,~,. ..

; .

".

,

,~

"

9

9, . t - , , ..

"

'

"

;'~'~

.

,'~.. ~ -.

'7"~

I~.,

~--.

-

"

'

"

,.-.'-=

.

.

.

~

C_.

'

.

..

.

"2;, ~ ' e

"

"K~ 9

", .

*. -

"

.

'

.

-

~

,~

".'

.

t

I

~-U

~"

~':

.

-

%,

I

I 9

'..

.-

::

J

k':- . ,,

r q~,

FIGURE 11 (top left)." Normal gastric m u c o s a of the pyIoric gland. Alph%-antitrypsin (AAT)-positive epithelial cells a r e found chiefly a t the b a s e of the glands. Mast cells ( a r r o w s ) also show a strong positive reaction for AAT. (Peroxidase-antiperoxidase technique, x 640.)

".,,

,,',,

:

.

.

9

-,<.

k ~"

-

,

",J

i/.'-

j

u

"

. y-s . ~

"

.Y

.~%r

\ q:N

FIGURE t 2 (top right). Pyloric m u c o s a a d j a c e n t to a gastric a d e n o c a r c i n o m a , M a n y AAT-positive cells of the pyloric glands a r e present a r o u n d t h e tumor, (Peroxid a s e - a n t i p e r o x i d a s e technique, x 368.) '

'

'

FIGURE 13 (/eft]. Brunner glands adjc~cent to a gastric a d e n o c a r c i n o m a , Alph%-antichymotrypsin-positive epithelial cells of the Brunner gTands a r e increased a r o u n d the tumor. (Peroxidase-antiperoxidase technique. x 368.)

stage in gastric carcinomas. T h e incidence o f tile three markers was higtler in advanced than in early cancers. Moreover, for advanced cancers A A T and AMG were detected more frequently in well-differ-

DISCUSSION

We have demonstrated that the expression o f AAT, ACT, and AMG differs by histologic type and 962

GASTRIC CARCINOMA (Tahara et al.)

entiated adenocarcinomas than in poorly differentiated adenocarcinomas. However, AMG was scattered or restricted to small areas o f the tumors. These results suggest that the presence of protease inhibitors in gastric carcinomas is related to the invasive growth of the tumors and that AAT may serve as a tissue tumor marker in well-differentiated adenocarcinomas of the stomach. Eighty per cent of the AAT-positive advanced well-differentiated adenocarcinomas were ACT-positive, and 40 per cent of these tumors were AMGpositive as well. These findings indicate that well-differentiated adenocarcinomas o f the stomach contain more than one of these protease inhibitors. However, the localization of A A T within a tumor was not necessarily the same as that o f A C T or AMG. It is well known that A A T is present in hepatoma inclusions containing AFP and AMG. 11-13 Moreover, Kodama et al. 14 reported that gastric adenocarcinomas with AFP-positive cells also contain two or three types of normal serum proteins, including AAT, albumin, and transferrin. Hbwever, in the present study AATpositive tumor cells and cytoplasmic inclusions were found to be AFP-negative. T h e incidence o f A C T within the tumor was also remarkably higher in advanced than in early cancers. In fact, serum A C T levels in patients ~(ith gastric carcinomas have been shown to increase with cancer stage35 Kelley et al. 16 also reported that serum A C T is useful in monitoring stomach cancers. Furthermore, Ogoshi et al. 15 reported the concentrations of A C T in the sera of patients with gastric cancers to be lower for signet ring cell carcinomas than for other carcinomas, but in the present study no remarkable difference in A C T expression by histologic type of gastric carcinoma was demonstrated. Furthermore, according to Takada et al., 17 A C T is present not only in the nuclei o f gastric carcinoma tumor cells but also in the nuclei o f lymphoid cells that have infiltrated the tumor tissue. In view of the enhancement of antibody formation by ACT, these authors suggested that A C T might act on lymphoid cells around the tumor cells; in the p r e s e n t study, however, we did not detect ACT in the lymphoid cells. In the normal gastric mucosa, A A T was detected chiefly in cardiac, pyloric, and Brunner's gland epithelial cells. T h e distribution of A A T was almost similar to that of lysozyme. 6 T h e number of AAT-positive epithelial cells, however, was much smaller than that of lysozyme-positive epithelial cells. These cells were localized frequently in the base of the glands. Although the presence of A A T in the small intestine has been reported, in the present study it was found neither in the intestinal metaplastic epithelium nor in the small intestine. In view of the intensity and distribution o f immunoreactive A A T in the stomach, both A A T and lysozyme appear to be synthesized by epithelial cells. Autodigestion by several proteolytic enzymes may physiologically protect the normal function and structural integrity of the gastric mucosa.

Kittas et al. 3 reported that the results of immunohistochemical staining o f gastric carcinomas for.lysozyme, ACT, and AAT differ by the site of origin of the tumors and that these differences reflect, in part, the distribution of AAT, ACT, and lysozyme in the normal gastric epithelium; we did not confirm these findings in the present study. In the present study gastric carcinomas arising from the fundus, in which A A T and A C T were either rare or absent, had numerous AAT- and ACT-positive cells. Thus, we speculate that in gastric tumors AAT and ACT, as well as lysozyme, are produced by t u m o r cells. Furthermore, the presence o f A A T in tumor cells appears to protect against potentially destructive proteases released by leukocytes a r o u n d t h e tumor. However, A A T immunoreactivity in gastric carcinomas showed no clear correlation to the degree of inflammatory cell infiltration. T h e numbers of AAT- and ACT-positive epithelial cells frequently increased in noncancerous mucosa adjacent to gastric carcinomas. Naito et al. 18 recently reported that cultured human gastric tumor cells produce plasnfinogen activator as well as tissue thromboplastin. M o r e o v e r , collagenase has frequently been demonstrated immunohistochemically in diffuse carcinomas of the stomach39 Therefore, focal increases in AAT- and ACT-containing epithelial cells around tumors may be considered a protective phenomenon against the high levels of thromboplastic or fibrinolytic activity and proteolytic enzymes o f gastric carcinomas. T h e presence of protease inhibitors in the tumor cells of gastric carcinomas raises the provoking question as to their effect on tumor growth or metastasis. Hatcher et al. 2~ reported that the decrease of surface protease activity is related to a decrease in proliferation o f the transformed cells. In the present study, however, the prognoses for patients with well-differentiated adenocarcinomas with A A T were apparently worse than those for patients with well-differentiated adenocarcinomas without AAT. T h e r e was no remarkable difference in the two-year survival rates between patients with ACT- or AMG-positiv'e and those with ACT- or AMG-negative well-differentiated adenocarcinomas. However, patients with ACT-negative poorly differentiated adenocarcinomas had extremely poor prognoses..These tumors might produce various proteases as well as o t h e r protein markers. T h e reasons for the severity of the malignancy o f well-differentiatedadenocarcinomas with A A T are unknown. It may be assumed that some of the well-differentiated adenocarcinomas with A A T also have proteases, as the cytoplasm of human mast cells contains both proteases and AAT. 21 Moreover, A A T has been d e m o n s t r a t e d to inhibit the blastogenic response of normal lymphocytes to phytohemagglutinin (PHA) n0nspecifically, z2-24 The immunosuppressive effect of A A T might be related to the biologic behavior. T h e relation, if any, b e t w e e n tumor cell A A T and prognosis in early gastric cancers remains to be elucidated. 963

HUMAN PATHOLOGY

Volume 15, No. 10 [October '1984]

REFERENCES 1. Geboes K, Ray MB, Rutgeerts P, et al: Morphological identification of alpha-l-antitrypsin in the human small intestine. Histopathology 6:55, 1982 2. Ray MB, Geboes K, Desrnet VJ: Demonstration of immunoreactive alpha-l-antitrypsin in gastric tumor. Hepatogastroenterology (suppl):89, 1980 3. Kittas C, Aroni K, Kotsls L, et al: Distribution of lysozyme, a xantichymotrypsin and a]-antitrypsin in adenocarcinomas of the stomach and large intestine. Virchows Arch [Pathol Anat] 398:139, 1982 4. Tahara E, h o H, Nakagami K, et al: Scirrhous argyrophil cell carcinoma of the stomach with multiple production of polypeptide hormones, amine, CEA, lysozyme, and hCG. Cancer 49:1904, 1982 5. Tahara E, Ito H, Shimamoto F, et ah Argyrophil cells ill early gastric carcinoma: an immunohistochemical and uhrastructural study. J Cancer Res Clin Oncol 103:187, 1982 6. Tahara E, ho H, Shimamoto F, et al: Lysozyme in human gastric carcinoma: a retrospective immunohistochenaical stud)'. Histopathology 6:409, 1982 7. Ito H, Tahara E: Human chorionic gonadotropin in human gastric carcinoma: a retrospective immunohistochemical stud)'. Acta PatholJpn 33:287, 1983 8. Sumiyoshi H, Taniyama K, Ito H, et al: Secretory component and immunoglobulin in human gastric carcinoma: an immunohistochemical study. Gann 74:166, 1984 9. Nagayo T: Histological classification of gastric cancer. In The General Rules for tile Gastric Cancer Study in Surgery and Pathology, Japanese Research Society for Gastric Cancer 1974, p 28 (in Japanese) 10. Sternberger LA: Immunocytochemistry, 2. New York', John Wiley & Sons, 1979, p 104 11. Palmer PE, Kostos CI, Wolfe HJ: Muhisystem fibrosis in an alpha-l-antitrypsin deficient (SZ) individhal (letter to the editor). Lancet 1:221, 1978 12. Palmer PE, Ucci AA, Wolfe HJ: Expression of protein markers in malignant hepatoma: evidence for genetic and epigenetic mechanisms. Cancer 45:1424, 1980 13. Wolfe HJ, Palmer PE: Alpha-l-antitrypsin: its immunohisto-

964

14.

15. 16. 17. 18. 19.

20.

21. 22. 23. 24.

chemical localization and significance in diagnostic pathology. In Steinberg SS (series ed): Masson Monographs in Diagnostic Pathology, vol 2. Delellis RA (ed): Diagnostic Immunohistochemistry. New York, Masson Publishing USA, 1981, p 227 Kodama T, Kameya T, Hirota T, et al: Production of alphafetoprotein, normal serum proteins, and human chorionic gonadotropin in stomach cancer: histologic and immunohistochemical analyses of 35 cases. Cancer 48:1674, 1981 Ogoshi K, Kondoh Y, Nakasaki H: Clinical evaluation of serum eq-antichymotrypsin, an acute phase reactant protein. Igakuno-ayumi 126:31, 1983 (in Japanese) Kelly VL, Cooper EH, Alexander C, et ah The assessment of a n t i c h y m o t r y p s i n in cancer m o n i t o r i n g . Biomedicine 28:209, 1978 Takada S, Tsuda M, Mitomi T, et al: Analysis of tile tissue and celhdar localization of alpha-1-antichymotrypsin by immunohistochemical teclmique. Gann 73:742, 1982 Naito S, Inoue S, Kinjo M, et al: Thromboplastic and fibrinolytic activities of cultured human gastric cancer cell lines. Gann 74:240, 1983 Ueda N; Isozaki H, Miyamoto M, et al: Studies on localization of collagenase in gastric cancer tissue by immunoperoxidase method. Proceedings of the Japanese Cancer Association, 41th Annual Meeting, Osaka, August 1982, p 318 (in Japanese) Hatcher VB, Oberman MS, Wertheim MS, et al: The relationship between surface protease activity and the rate of cell proliferation in normal and transformed cells. Biochem Biophys Res Commun 76:602, 1977 Benitez-Bribiesca L, Freyre R, Vega G: Alpha-l-antitrypsm in human mast cells: immunofluorescent localization. Life Sci 13:631, 1973 Arora PK, Miller ItC: eq-Antitrypsin is an effector of immunological stasis. Nature 274:589, 1978 Samsk R, Edelstein R, Israel L: Immunosuppresslve effect of acute phase reactant proteins in vitro and its relevance to cancer. Cancer Immunol Immunother 13:38, 1982 Nakanishi K: Experimental and clinical studies of serum immunosuppressive factors in t u m o r bearing host. Med J Hiroshima Univ 31:855, 1983 (in Japanese)