Mucosal-cell Counts in Ulcerative and Granulomatous Colitis S H E L D O N C. SOMMERS, M.D.,
AND B U R T O N I. K O R E L I T Z ,
M.D.
Departments of Pathology and Medicine, Lenox Hill Hospital, 100 East 77th Street, New York, New York 10021
ABSTRACT
CELLS in rectal biopsies as a means of evaluating the inflammatory reactions in ulcerative colitis and Crohn's disease may help to distinguish between these two conditions and to identify their responses to therapy. 3,4 A brief review of the technics used, the results obtained, a n d some f u r t h e r applications of COUNTING
Received September 9, 1974; accepted for publication October 3, 1974. Presented in part at the Spring meeting of ASCP, Los Angeles, February 1974. Address reprints to Dr. Sommers, Director of Laboratories, Lenox Hill Hospital, 100 East 77th Street, New York, New York 10021. 359
mucosal-cell counts is presented as a way of encouraging their wider employment. Materials and Methods Biopsies of the rectum or rectosigmoid were obtained with cutting forceps from patients with ulcerative colitis or granulomatous colitis. Specimens included the mucosa, t h e muscularis mucosae, and often some submucosa. Tissues were fixed in Bouin's solution, processed in a routine manner to prepare paraffin-embedded sections, and stained
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
Sommers, Sheldon C , and Korelitz, Burton I.: Mucosal-cell counts in ulcerative and granulomatous colitis. Am J Clin Pathol 63: 3 5 9 - 3 6 5 , 1975. Epithelial and connective-tissue cells were counted in rectal mucosal biopsies from 215 patients with ulcerative colitis, 98 patients with granulomatous colitis, and 50 controls. T h e results were analyzed statistically. Significantly decreased mucous goblet cells were found both in sigmoidoscopically abnormal ulcerative colitis and in granulomatous colitis, and they increased during the healing process. More pyknotic and karyorrhectic epithelial cells occurred in active ulcerative colitis than in granulomatous colitis. Inactive ulcerative colitis still manifested histologic evidence of acute and chronic inflammation, while sigmoidoscopically normal granulomatous colitis biopsies after previous gross rectal disease showed significantly increased macrophages in the lamina propria. Cell counts were valuable for differential diagnosis after the sigmoidoscopic appearance became normal. The acute inflammation of ulcerative colitis, as indicated by neutrophils, was d e c r e a s e d most notably following t h e r a p y with p r e d n i s o n e or 6-mercaptopurine. Chronic inflammation associated with fewer plasma cells was decreased after salicylazosulfapyridine as well as either of the other two drugs; macrophages, indicators of healing, increased most after 6-mercaptopurine combined with another anti-inflammatory agent. (Key words: Ulcerative colitis; Granulomatous colitis; Mucosal-cell counts.)
360
SOMMERS AND KORELITZ
A histopathologic diagnosis has been rendered for each specimen, but this was rarely useful in discrimination between ulcerative and granulomatous colitis.3 Independently, and without other clinical or pathologic information, cell counts were performed. Crypt epithelial cells were counted at the gland bases to a total of 500 cells, including mucous goblet, nonmucous, mitotic, pyknotic and karyorrhectic (fragmented nucleated) cells, a total of five categories. Also, 500 connective-tissue cells of the lamina propria of the same area, comprising fibroblasts, macrophages, mast cells, lymphocytes, plasma cells, eosinophils, and neutrophils, were counted. Argentaffin and Paneth cells have not been separately identified. No basophil leukocytes were observed. T o avoid counting overlapping
fields, when one section was insufficient, several serial sections were skipped before continuing the count. With an ordinary laboratory counter, the entire procedure took about 15 minutes per biopsy specimen. The epithelial and connective-tissue cell counts were separately calculated as percentages of individual cell types. After the counting was completed, the clinical groups of cases were assembled for purposes of clinical-pathologic correlation. Means and standard deviations were calculated. T h e proportional area in the mucosa composed of connective tissue was determined. Fifteen each of biopsies from controls and patients with untreated ulcerative colitis were selected for this purpose. The selection was determined by the normal distribution of cells counted per field. These included patients with counts at the group mean (5), one standard deviation above and below the mean (6), and two standard deviations above and below the mean (4). Six fields were mapped per specimen by projecting images onto cardboard and outlining the epithelial glands in pencil. T h e cardboard circles were thereafter cut u p into the epithelial and connective-tissue portions, and these were separately weighed for each case. T h e n the 15 control and 15 ulcerative colitis biopsies were analyzed for mean percentages of area of epithelium and connective tissue. Cell populations per complete field of lamina propria could therefore be determined, and results were used as correction factors for connectivetissue cells per field found in the original counts. Although connective-tissue cells could be calculated per unit area, such as per 100 or 5,000 square micra, 1,2 the results would depend partly upon section thickness. T h e thickness of the sections was 8 - 1 0 micra in previous studies; it was 5 - 6 micra in the present material. It was
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
with the trichrome hematoxylin-phloxinesafranin method. Other fixatives and polychrome stains such as Giemsa that color mast cells and eosinophil granules were equally useful. T h e total biopsies obtained exceeded 400, and in this report biopsies from 363 individuals are analyzed. T h e diagnostic categories assigned were based on clinical history, gastrointestinal x-rays, and results of sigmoidoscopic examinations. T h e patients were divided into those whose sigmoidoscopic findings were abnormal (positive, +) or normal (negative, - ) . There were 160 patients who had sigmoidoscopically positive ulcerative colitis, including 43 who were untreated and 117 who were medically treated, and 55 patients who had sigmoidoscopically negative treated ulcerative colitis; 44 patients had granulomatous colitis that was sigmoidoscopically positive and 54, sigmoidoscopically negative granulomatous colitis. Fifty patients who did not have colitis served as controls; they had had biopsies because of diarrhea or other gastrointestinal symptoms not caused by inflammatory intestinal disease.
A.J.C.P. —Vol. 63
March 1975
361
CELL COUNTS IN COLITIS
Table 1. Epithelial Cells in Rectal Biopsies of Patients with Ulcerative and Granulomatous Colitis and Controls (Means ± S.D.) No. of Patients
Mucous Goblet
Nonmucous
Mitotic
Pyknotic
Fragmented Nuclei
%
%
%
%
%
43
32.9 ± 16.1
51.1 ± 14.7
2.55 ± 1.55
5.80 ±3.96
1.55 ± 1.62
Treated ulcerative colitis, sigmoidoscopy +
117
39.8 ± 17.6
47.7 ± 17.3
2.20 ± 1.55
9.33 ±4.58
1.10 ± 1.23
Treated ulcerative colitis, sigmoidoscopy-
55
±
2.24 ± 1.69
8.26 ±4.83
0.144 ±0.211
Granulomatous colitis, sigmoidoscopy+
44
43.9 ± 15.9
2.21 ± 1.09
4.99 ±2.20
0.669 ±0.814
Granulomatous colitis, sigmoidoscopy-
54
±
50.6 7.76
±
41.7 7.66
3.39 ±7.99
5.99 ±3.76
0.360 ± 0.460
Controls, sigmoidoscopy—
50
±
49.7 8.92
±
41.5 8.05
2.23 ± 1.78
6.01 ±3.01
0.262 ±0.431
53.5 6.55
±
35.9 8.01
48.2 ± 16.1
Mucous goblet cells were significandy reduced in untreated or treated sigmoidoscopically positive ulcerative colitis, compared with either treated sigmoidoscopically negative ulcerative colitis or controls (p < 0.001). Also, mucous goblet cells were reduced in sigmoidoscopically positive granulomatous colitis, compared with sigmoidoscopically negative granulomatous colitis (p < 0.01) or controls (p < 0.05). Fragmented epithelial nuclei were more numerous in untreated ulcerative colitis than in treated sigmoidoscopically negative or control groups (both p < 0.001). In sigmoidoscopically positive granulomatous colitis there also were more fragmented nuclei than in either the sigmoidoscopically negative (p < 0.025) or control groups (p < 0.005). As already reported, 1 treated ulcerative Results colitis that was sigmoidoscopically positive In Table 1 are given the mean percent- had more pyknotic epithelial cells than ages and standard deviations of epithelial untreated ulcerative colitis, granulomacells in rectal biopsies of patients with tous colitis, or controls (all p < 0.001). ulcerative and granulomatous colitis, each Finally, in sigmoidoscopically positive uldivided into sigmoidoscopically positive cerative colitis, whether untreated or and negative groups for analysis, and in treated, more fragmented nuclei were rectal biopsies of normal controls. found than in sigmoidoscopically positive
considered best in view of this variation to report cells per field; each field was 38,000 square micra. T h e evaluation of epithelial cells was based on percentages of the total cells counted. Statistical comparisons were made by t tests. Replicability of the method was tested by comparing 17 pairs of counts made inadvertently. T h e areas analyzed were not necessarily identical on repeated counting. Cell types that made u p 7% or more of the epithelium or connective tissue differed by an average of 2.25%. Minority cell types that comprised less than 7% of the total differed by an average of 24.8% when the same specimen was counted two or three times.
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
Untreated ulcerative colitis, sigmoidoscopy+
362 lis
SOMMERS AND KORELITZ
a
CO
CM
O)
o
in
00 CO
to
q
CM Tf
00
00
to
+1
+1
+1
o
o
o
— +1 -^
CO CM
+1
m
— 2 0
r^
h
o
itroph
(/J f=
a
d
00
—
00
CM
o
ai
o
a>
o
to
o d
CO 00
to
to r-^
in r»>
iri
CO
+1
+1
+1
o
00 Ol
2
•
O) to
*
"+1
o
a> CO
o
CM CO
CO
in
d
d
o
d +1
CM
—< d
+1
+1
o
O
CM
CTi
d
d
a; --i
r~
-^ +1 •
*
o
+1
+1
m m 00
o
d
•
*
^ o
d
o o>
d
CM CM
CM
to
to
Tt"
CM
00
— +1
+1
— +1
— — +1
to
to
CM
in
Tf
•
Tj-
d Tf
t^ CO
d CO
m o
Tt
O
o
m
-^ •9-
o m
to
-Jr^
•«•
o
o m CM
tCe
d +1
00 CO
CO CM
ty> Tf CM
o m CM
d
d
d
d
d
o o
CM CM
Tf
CM CO
CM
CO
CO
CO
CM
+1
+1
+1
pha ges
d +1
—i
+1
o u
ON 00
a>
00
00
•
*
*
s
,
o m
E
CO
C
D
to
00
t^
l>
Tf 00
00 •
d +1
T^
+1
+1
00
in
00
d CM
t^
d
— d ^
+1
to to
00
—
•o '5 > '5 E
E
S .bp _u bp JJ bp 4j
;5 tj
V
u
H
3
in
w
;s
in
;« "o
in
V
00
CO
+1
+ >-*
+1 CM in
to
in
-^
o d
to
o
+1
3 T3 u
00
—
3 T3 '5 3
m
r>^
> 00
o
t^
to
—* +
t m
d +1
•
d +1
CO
— i
+1
CO
•* CM
• * .
d +1
CO
0
+1
^
o
«
iri
+1
CM
— d
2
i>
CO CM
l>
ra
to
CO CM
d +1
5
"*< ^_
CO
—
-^
*
«
(/)
:s 1 "5 + >-, a. >-* 3 a. §o 0 0 •3 u
V
CM
to
ti«
d 00
— +1
—
to
AND B U R T O N I. K O R E L I T Z ,
M.D.
Departments of Pathology and Medicine, Lenox Hill Hospital, 100 East 77th Street, New York, New York 10021
ABSTRACT
CELLS in rectal biopsies as a means of evaluating the inflammatory reactions in ulcerative colitis and Crohn's disease may help to distinguish between these two conditions and to identify their responses to therapy. 3,4 A brief review of the technics used, the results obtained, a n d some f u r t h e r applications of COUNTING
Received September 9, 1974; accepted for publication October 3, 1974. Presented in part at the Spring meeting of ASCP, Los Angeles, February 1974. Address reprints to Dr. Sommers, Director of Laboratories, Lenox Hill Hospital, 100 East 77th Street, New York, New York 10021. 359
mucosal-cell counts is presented as a way of encouraging their wider employment. Materials and Methods Biopsies of the rectum or rectosigmoid were obtained with cutting forceps from patients with ulcerative colitis or granulomatous colitis. Specimens included the mucosa, t h e muscularis mucosae, and often some submucosa. Tissues were fixed in Bouin's solution, processed in a routine manner to prepare paraffin-embedded sections, and stained
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
Sommers, Sheldon C , and Korelitz, Burton I.: Mucosal-cell counts in ulcerative and granulomatous colitis. Am J Clin Pathol 63: 3 5 9 - 3 6 5 , 1975. Epithelial and connective-tissue cells were counted in rectal mucosal biopsies from 215 patients with ulcerative colitis, 98 patients with granulomatous colitis, and 50 controls. T h e results were analyzed statistically. Significantly decreased mucous goblet cells were found both in sigmoidoscopically abnormal ulcerative colitis and in granulomatous colitis, and they increased during the healing process. More pyknotic and karyorrhectic epithelial cells occurred in active ulcerative colitis than in granulomatous colitis. Inactive ulcerative colitis still manifested histologic evidence of acute and chronic inflammation, while sigmoidoscopically normal granulomatous colitis biopsies after previous gross rectal disease showed significantly increased macrophages in the lamina propria. Cell counts were valuable for differential diagnosis after the sigmoidoscopic appearance became normal. The acute inflammation of ulcerative colitis, as indicated by neutrophils, was d e c r e a s e d most notably following t h e r a p y with p r e d n i s o n e or 6-mercaptopurine. Chronic inflammation associated with fewer plasma cells was decreased after salicylazosulfapyridine as well as either of the other two drugs; macrophages, indicators of healing, increased most after 6-mercaptopurine combined with another anti-inflammatory agent. (Key words: Ulcerative colitis; Granulomatous colitis; Mucosal-cell counts.)
360
SOMMERS AND KORELITZ
A histopathologic diagnosis has been rendered for each specimen, but this was rarely useful in discrimination between ulcerative and granulomatous colitis.3 Independently, and without other clinical or pathologic information, cell counts were performed. Crypt epithelial cells were counted at the gland bases to a total of 500 cells, including mucous goblet, nonmucous, mitotic, pyknotic and karyorrhectic (fragmented nucleated) cells, a total of five categories. Also, 500 connective-tissue cells of the lamina propria of the same area, comprising fibroblasts, macrophages, mast cells, lymphocytes, plasma cells, eosinophils, and neutrophils, were counted. Argentaffin and Paneth cells have not been separately identified. No basophil leukocytes were observed. T o avoid counting overlapping
fields, when one section was insufficient, several serial sections were skipped before continuing the count. With an ordinary laboratory counter, the entire procedure took about 15 minutes per biopsy specimen. The epithelial and connective-tissue cell counts were separately calculated as percentages of individual cell types. After the counting was completed, the clinical groups of cases were assembled for purposes of clinical-pathologic correlation. Means and standard deviations were calculated. T h e proportional area in the mucosa composed of connective tissue was determined. Fifteen each of biopsies from controls and patients with untreated ulcerative colitis were selected for this purpose. The selection was determined by the normal distribution of cells counted per field. These included patients with counts at the group mean (5), one standard deviation above and below the mean (6), and two standard deviations above and below the mean (4). Six fields were mapped per specimen by projecting images onto cardboard and outlining the epithelial glands in pencil. T h e cardboard circles were thereafter cut u p into the epithelial and connective-tissue portions, and these were separately weighed for each case. T h e n the 15 control and 15 ulcerative colitis biopsies were analyzed for mean percentages of area of epithelium and connective tissue. Cell populations per complete field of lamina propria could therefore be determined, and results were used as correction factors for connectivetissue cells per field found in the original counts. Although connective-tissue cells could be calculated per unit area, such as per 100 or 5,000 square micra, 1,2 the results would depend partly upon section thickness. T h e thickness of the sections was 8 - 1 0 micra in previous studies; it was 5 - 6 micra in the present material. It was
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
with the trichrome hematoxylin-phloxinesafranin method. Other fixatives and polychrome stains such as Giemsa that color mast cells and eosinophil granules were equally useful. T h e total biopsies obtained exceeded 400, and in this report biopsies from 363 individuals are analyzed. T h e diagnostic categories assigned were based on clinical history, gastrointestinal x-rays, and results of sigmoidoscopic examinations. T h e patients were divided into those whose sigmoidoscopic findings were abnormal (positive, +) or normal (negative, - ) . There were 160 patients who had sigmoidoscopically positive ulcerative colitis, including 43 who were untreated and 117 who were medically treated, and 55 patients who had sigmoidoscopically negative treated ulcerative colitis; 44 patients had granulomatous colitis that was sigmoidoscopically positive and 54, sigmoidoscopically negative granulomatous colitis. Fifty patients who did not have colitis served as controls; they had had biopsies because of diarrhea or other gastrointestinal symptoms not caused by inflammatory intestinal disease.
A.J.C.P. —Vol. 63
March 1975
361
CELL COUNTS IN COLITIS
Table 1. Epithelial Cells in Rectal Biopsies of Patients with Ulcerative and Granulomatous Colitis and Controls (Means ± S.D.) No. of Patients
Mucous Goblet
Nonmucous
Mitotic
Pyknotic
Fragmented Nuclei
%
%
%
%
%
43
32.9 ± 16.1
51.1 ± 14.7
2.55 ± 1.55
5.80 ±3.96
1.55 ± 1.62
Treated ulcerative colitis, sigmoidoscopy +
117
39.8 ± 17.6
47.7 ± 17.3
2.20 ± 1.55
9.33 ±4.58
1.10 ± 1.23
Treated ulcerative colitis, sigmoidoscopy-
55
±
2.24 ± 1.69
8.26 ±4.83
0.144 ±0.211
Granulomatous colitis, sigmoidoscopy+
44
43.9 ± 15.9
2.21 ± 1.09
4.99 ±2.20
0.669 ±0.814
Granulomatous colitis, sigmoidoscopy-
54
±
50.6 7.76
±
41.7 7.66
3.39 ±7.99
5.99 ±3.76
0.360 ± 0.460
Controls, sigmoidoscopy—
50
±
49.7 8.92
±
41.5 8.05
2.23 ± 1.78
6.01 ±3.01
0.262 ±0.431
53.5 6.55
±
35.9 8.01
48.2 ± 16.1
Mucous goblet cells were significandy reduced in untreated or treated sigmoidoscopically positive ulcerative colitis, compared with either treated sigmoidoscopically negative ulcerative colitis or controls (p < 0.001). Also, mucous goblet cells were reduced in sigmoidoscopically positive granulomatous colitis, compared with sigmoidoscopically negative granulomatous colitis (p < 0.01) or controls (p < 0.05). Fragmented epithelial nuclei were more numerous in untreated ulcerative colitis than in treated sigmoidoscopically negative or control groups (both p < 0.001). In sigmoidoscopically positive granulomatous colitis there also were more fragmented nuclei than in either the sigmoidoscopically negative (p < 0.025) or control groups (p < 0.005). As already reported, 1 treated ulcerative Results colitis that was sigmoidoscopically positive In Table 1 are given the mean percent- had more pyknotic epithelial cells than ages and standard deviations of epithelial untreated ulcerative colitis, granulomacells in rectal biopsies of patients with tous colitis, or controls (all p < 0.001). ulcerative and granulomatous colitis, each Finally, in sigmoidoscopically positive uldivided into sigmoidoscopically positive cerative colitis, whether untreated or and negative groups for analysis, and in treated, more fragmented nuclei were rectal biopsies of normal controls. found than in sigmoidoscopically positive
considered best in view of this variation to report cells per field; each field was 38,000 square micra. T h e evaluation of epithelial cells was based on percentages of the total cells counted. Statistical comparisons were made by t tests. Replicability of the method was tested by comparing 17 pairs of counts made inadvertently. T h e areas analyzed were not necessarily identical on repeated counting. Cell types that made u p 7% or more of the epithelium or connective tissue differed by an average of 2.25%. Minority cell types that comprised less than 7% of the total differed by an average of 24.8% when the same specimen was counted two or three times.
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
Untreated ulcerative colitis, sigmoidoscopy+
362 lis
SOMMERS AND KORELITZ
a
CO
CM
O)
o
in
00 CO
to
q
CM Tf
00
00
to
+1
+1
+1
o
o
o
— +1 -^
CO CM
+1
m
— 2 0
r^
h
o
itroph
(/J f=
a
d
00
—
00
CM
o
ai
o
a>
o
to
o d
CO 00
to
to r-^
in r»>
iri
CO
+1
+1
+1
o
00 Ol
2
•
O) to
*
"+1
o
a> CO
o
CM CO
CO
in
d
d
o
d +1
CM
—< d
+1
+1
o
O
CM
CTi
d
d
a; --i
r~
-^ +1 •
*
o
+1
+1
m m 00
o
d
•
*
^ o
d
o o>
d
CM CM
CM
to
to
Tt"
CM
00
— +1
+1
— +1
— — +1
to
to
CM
in
Tf
•
Tj-
d Tf
t^ CO
d CO
m o
Tt
O
o
m
-^ •9-
o m
to
-Jr^
•«•
o
o m CM
tCe
d +1
00 CO
CO CM
ty> Tf CM
o m CM
d
d
d
d
d
o o
CM CM
Tf
CM CO
CM
CO
CO
CO
CM
+1
+1
+1
pha ges
d +1
—i
+1
o u
ON 00
a>
00
00
•
*
*
s
,
o m
E
CO
C
D
to
00
t^
l>
Tf 00
00 •
d +1
T^
+1
+1
00
in
00
d CM
t^
d
— d ^
+1
to to
00
—
•o '5 > '5 E
E
S .bp _u bp JJ bp 4j
;5 tj
V
u
H
3
in
w
;s
in
;« "o
in
V
00
CO
+1
+ >-*
+1 CM in
to
in
-^
o d
to
o
+1
3 T3 u
00
—
3 T3 '5 3
m
r>^
> 00
o
t^
to
—* +
t m
d +1
•
d +1
CO
— i
+1
CO
•* CM
• * .
d +1
CO
0
+1
^
o
«
iri
+1
CM
— d
2
i>
CO CM
l>
ra
to
CO CM
d +1
5
"*< ^_
CO
—
-^
*
«
(/)
:s 1 "5 + >-, a. >-* 3 a. §o 0 0 •3 u
V
CM
to
ti«
d 00
— +1
—
to
