GASTROENTEROLOGY
1992;103:1790-1796
Subepithelial Collagen Table Thickness in Colon Specimens From Patients With Microscopic Colitis and Collagenous Colitis EDWARD LEE, LAWRENCE R. SCHILLER, DORIS VENDRELL, CAROL A. SANTA ANA, and JOHN S. FORDTRAN Departments of Internal Medicine and Pathology, Baylor University Medical Center; Department of Pathology, Veterans Administration Medical Center; and University of Texas Southwestern Medical Center, Dallas, Texas
Microscopic colitis and collagenous colitis are similar conditions that are differentiated by the presence or absence of subepithelial collagen table thickening. To better understand the relationship between these two disorders and the role of collagen table thickening in the pathogenesis of diarrhea, colonic mucosal biopsy specimens from 24 patients with microscopic or collagenous colitis and 9 control subjects were analyzed using a computerassisted morphometric method to evaluate the average thickness of the subepithelial collagen table. The collagen table thickness in colitis patients taken together formed a multimodal rather than a unimodal distribution. There was no tendency for collagen table thickening to increase with age or with duration of symptoms. In general, the types and distribution of inflammatory cells were similar in patients with normal and thickened collagen tables. Stool weight correlated with lamina propria cellularity but not with collagen table thickening. The multimodal distribution of collagen table thickening and the lack of correlation with age, duration of symptoms, or inflammation suggest that microscopic colitis and collagenous, colitis are disalthough the inflammatory crete conditions, changes in the two conditions are similar. Moreover, because stool weight correlates with lamina propria cellularity but not with collagen table thickening, diarrhea probably is caused by the inflammatory changes and not by collagen table thickening per se.
I
n the past 15 years, two new forms of inflammatory bowel disease, microscopic colitis and collagenous colitis, have been described.‘-25 These conditions have many features in common. Both present with watery diarrhea caused by reduced colonic fluid absorption. In both conditions the colon appears normal by barium enema and colonoscopy, but biopsies of the colon show that the lamina propria
contains an inflammatory infiltrate of plasma cells and neutrophils and the epithelium is invaded by lymphocytes and occasional neutrophils. The two conditions are separated histologically only by the presence or absence of subepithelial collagen table thickening. If the subepithelial collagen table is considered abnormally thick, the disorder is labeled collagenous colitis; if not, the disorder is labeled microscopic (or lymphocytic) colitis. However, it is not always possible to determine by routine microscopy whether the collagen layer is thickened, and even expert pathologists may disagree as to whether collagenous colitis or microscopic colitis is present in a given case.” With so many similarities, it has been suggested that these two conditions are variants of the same process with a spectrum of collagen table thickening, perhaps related to age, duration of illness, or severity of inflammation.lg Whether these histological patterns represent different responses to the same disease or two discrete disease processes, the pathophysiological importance of the thickened collagen table is uncertain; some believe that it forms a barrier that inhibits fluid absorption by the colon,’ while others contend that mucosal inflammation is the chief determinant of colonic fluid malabsorption.6 To gain insight into these issues, we carefully measured subepithelial collagen table thickness in the colon and rectum of patients with diagnoses of microscopic or collagenous colitis. Measurements were performed using a novel computer-assisted morphometric method that minimizes observer bias and allows the average collagen table thickness for each patient to be calculated. All biopsy specimens were also evaluated blindly by a previously described computer-assisted morphometric analysis to enumerate the severity of inflammation and the types of inflammatory cells in the lamina propria.” 0 1992
by the American Gastroenterological 0016-5085/92/$3.00
Association
COLLAGEN TABLE IN MICROSCOPIC/COLLAGENOUS COLITIS
December 19%
With knowledge of the average collagen table thickness in each patient, we were then able to address five issues concerning these disorders. First, we plotted collagen table thickness as a frequency distribution, which enabled us to determine whether collagen table thickness in these patients had a unimodal or multimodal distribution.2628 A unimodal, normal distribution would be consistent with the hypothesis of a single disease process; a multimodal distribution would be consistent with either separate disease processes or an extraneous influence that might influence collagen table thickness, such as age, duration of disease, or degree of inflammation. Second, we plotted collagen table thickness as a function of age; a positive correlation would suggest that collagen table thickening is related to aging. Third, we plotted collagen table thickness as a function of duration of illness: a positive correlation would suggest that collagen table thickening is a secondary response to the primary disease rather than an inherent characteristic. Fourth, we plotted collagen table thickness as a function of the severity of mucosal inflammation; a positive correlation would suggest that collagen thickness is determined in part by the degree to which the lamina propria is inflamed. And fifth, we plotted collagen table thickness as a function of daily stool weight; a positive correlation would suggest that collagen table thickening contributes to colonic fluid malabsorption. Materials and Methods Patients and Subjects Demographic and clinical information on 24 patients with clinical pathological diagnoses of microscopic colitis or collagenous colitis is shown in Table 1. Each of these patients had chronic diarrhea (2-348 months’ duration) with stool weights ranging from 246 to 1438 g/24 h. In each case the colonic mucosa was normal when inspected with the colonoscope, but mucosal biopsies were interpreted as indicating microscopic or collagenous colitis. A control group was composed of four normal volunteers who answered classified advertisements and five patients undergoing colonoscopy for evaluation of rectal bleeding or for follow-up after polypectomy. This group included five men and four women aged 37-70 years. Subjects gave written informed consent before participating in this study, which had been approved by the Institutional Review Board for Human Protection. In each subject, colonoscopy was negative and biopsy results were interpreted as normal. Some results from this group have been reported previously.6 Biopsy
Technique
and Processing
Participants were prepared for colonoscopy using Golytely lavage (Braintree Laboratories, Inc., Braintree, MA). When possible, biopsy specimens were obtained, us-
1791
ing Olympus forceps (Olympus Corporation of America, North Chicago, IL), from six regions of the colon: cecum, ascending colon, transverse colon, descending colon, sigmoid colon, and rectum. In some individuals, biopsy specimens were not obtained from all regions. Specimens were fixed in Carson’s buffered formalin, embedded in paraffin, and cut into 6-pm sections on a Leitz 1512 microtome (Australian Biomedical Corp., Melbourne). Sections were stained with H&E or Masson’s trichrome stain. Slides were reviewed by one pathologist (D.V.) who rendered a clinical pathological diagnosis of either microscopic colitis or collagenous colitis and were then coded in such a way that the second pathologist (E.L.), who conducted the morphometric analysis, would be unaware of the identity of the patient or subject, from which group they came (i.e., normal, microscopic colitis, or collagenous colitis), and from which part of the colon the specimens were obtained.
Morphometric
Analysis
A computer-assisted method was developed to measure the thickness of the subepitheiial collagen table without bias to the thickest parts. The &X-magnified image of a trichrome-stained biopsy specimen was conveyed through a Zeiss photomicroscope (Carl Zeiss, Inc., Thornwood, NY) equipped with a Zeiss CCTV downtube adapter to a Hitachi KP-ClOOAcolor video camera (Hitachi Denshi, Ltd., Tokyo, Japan) and was displayed on a CPD1290 RGB color video monitor (Sony, Fujisaua, Japan). A Videometric 150 image analyzer (American Sunovisior, San Diego, CA) was calibrated to the aox image using a stage micrometer and was used to measure the thickness of the subepithelial collagen table. Measurements of collagen table thickness were made at aoym intervals across the surface of well-oriented colon biopsy specimens, so that even if there was an unconscious bias to the initial placement of the specimen for measurement the remaining measurements would be made in an unbiased fashion. A total of 20 measurements were made per region of the colon. In some patients well-oriented sections were not available from all regions. Collagen table thicknesses measured in each region of the colon were averaged together for purposes of statistical analysis. Quantitative morphometric analysis of biopsy specimens to assess inflammation was carried out using a computer-assisted point-counting technique previously validated and described in detail.6 In brief, a dot from a video screen was projected over the magnified image of a welloriented biopsy specimen, and the structure on which the dot landed was identified. The dot was then advanced in a rectilinear pattern across the specimen until 98 points were identified. Other regions of the same biopsy specimen were then evaluated, and the process was repeated until seven areas were counted per specimen from each of the six regions of the colon. More than 4000 points were identified per patient. In some patients, well-oriented sections were not available from all regions of the colon. The results were expressed as the percentage of dots landing on a given structure; this reflects the relative area encompassed by that structure.
1792
GASTROENTEROLOGY
LEE ET AL.
ofMorphometric
Table I. Selected Results
Patient
A@ (Yrl
SEX
1
57
F
2
61
M
Duration of diarrhea WI
Stool weight (S/24 hl
2
1078
30
509
3
50
F
3
564
4
72
F
60
437
5
77
M
2
493
6
55
F
12
317
7
56
F
7
1028
8
58
F
11
1269
9
43
F
84
316
348
429
3
246
10
61
F
11
68
M
12
40
F
192
349
13
69
F
156
1105
14
50
F
156
431
15
42
F
18
424
16
41
F
54
451
17
52
F
5
516
18
70
F
8
600
19
65
F
11
408
20
51
F
240
661
21
44
F
30
573
22
55
F
3
899
23
84
F
54
1438
24
59
F
11
401
Analysis
Clinical pathological diagnosis”
Confidence intervals for control group
MC MC cc MC MC MC MC MC MC MC cc MC MC cc MC MC MC cc cc cc cc cc cc MC
Vol. 103, No. 6
in Individual Patients With Microscopic or Collagenous Colitis Lamina propria cellularity
Neutropbils
Plasma cells
Lymphocytes
MaCUDphases
Fibroblasts
Eosinophils
C&Se,, table thickness
(couilts per 100)
WI
Cm1
7d
15
14b
3
3
6
2.41
7ob
16
13b
3
3
6
2.47
59b
14
6b
3
3
2.68
67b
16
2
7b
4
3
2.76
9
0
3
2
5
3.10
58b
12
2
2
3
6
3.16
68b
18b
5
4
3
2
3.35
68b
16
4
4
3
5
3.54
Ssb
18
4
2
2
5
4.44b
48
61b
11
9b
4
3
5
4.71b
50
12
2
3
3
5
5.07b
72b
19b
5
2
3
3
5.24b
72b
19b
lob
3
2
5
5.88b
66b
15
7b
4
2
4
6.44b
75h
17b
13b
2
2
6
8.66b
63*
16
4
3
3
4
6.98b
74b
17b
10b
3
2
6
7.05b
65b
13
4
4
3
4
7.06b
6gb
19b
3
3
2
3
9.48b
65b
9
0
3
3
7
12.97b
83b
15
2
3
2
3
13.03b
71b
16
3
2
3
4
13.51b
82b
22b
7b
2
2
17.75b
68b
23b
5
3
2
4
37-57
8-16
o-5
O-4
o-4
3-7
18.50b
o-7
1.89-3.65
“MC. microscopic colitis; CC, collaganous colitis. bAbnormal as defined by confidence intervals in control group.
Statistics Data were analyzed by x2 analysis, analysis of variance (ANOVA), group t test, and Pearson’s coefficient of correlation using a computer program (StatPak Gold; Walonick Associates, Inc., Minneapolis, MN).2~2QProbability values of ~0.05 were considered statistically significant. Confidence intervals were calculated as means + ZSD.
Results Collagen table thickness varied little throughout the colon in normal subjects. Collagen table thickness in the cecum, ascending colon, transverse colon, descending colon, sigmoid colon, and rectum averaged 2.83, 2.67, 2.73, 2.75, 2.53, and 3.46 pm, respectively. The average thickness in the rectum was slightly greater than elsewhere in the colon by ANOVA (P < 0.05). For purposes of defining a normal range, all measurements in the normal subjects were the upper limit of normal averaged together; (mean + 2SD) for collagen table thickness was 3.65 pm. Results of measurement of collagen table thickness and morphometric analysis for each patient are shown in Table 1. Patients are arranged in order of average collagen table thickness. For comparison, 95% confidence intervals derived from the control group are also shown. Collagen table thickness was abnormal in 16 of 24 patients; all but 1 of these were women. Of the 8 remaining patients with normal collagen table thickness, 2 were men and 6 were
women. This difference in sex ratio was not statistically significant by x2 analysis. As shown in Figure 1, a frequency polygon of collagen table thickness showed a multimodal distribution, with some colitis patients in the normal range (~3.65 pm) and others with collagen table thicknesses greater than normal. The Kolmogorov-Smirnov statistic for normality was 1.24, strongly suggesting that collagen table thickness in the colitis 701
-fi % 30 r3 B
I I\
Normal Subjects
Colitis Patients
0
2.0
4.0
6.0
8.0 10.0 12.0 CT Thickness @I)
14.0
16.0
18.0
Figure 1. Frequency polygons for percentage of normal subjects or colitis patients with different collagen table thicknesses. The normal subjects form a unimodal population, whereas the colitis patients are a mutimodal population.
December
1992
COLLAGEN
patients was not normally distributed (P < 0.01). This statistic was almost as large (1.19) when only biopsy results from women were considered (P < 0.01). As shown in Figure 2 and Table 2, there was no significant correlation of collagen table thickness with age, duration of diarrhea, or lamina propria cellularity (an index of the overall degree of inflammation). Unadjusted P values for correlation coefficients suggested that collagen table thickness correlated positively with the number of plasma cells and inversely with the numbers of lymphocytes and macrophages; however, these correlations were not statistically significant when Bonferroni adjustment for multiple comparisons was applied.” There was no significant correlation with numbers of eosinophils, fibroblasts, or neutrophils (Table 2). Comparison of morphometric analysis in colitis patients with normal collagen table thickness (~3.65
.
.
. .
lo% .
5-
0’
??
.
40
“30
50
;
. .
-..
60 Age (yr)
.
80
1793
Table 2. Correlations Of Collagen Table Thickness With Clinical Data And Morphometric Results r Clinical data Age Duration of diarrhea Stool weight Morphometric results Lamina propria cellularity Plasma cells Lymphocytes Macrophages Eosinophils Fibroblasts Neutrophils
Unadjusted
0.09 0.02 0.16
NS NS NS
0.33 0.42 -0.42 -0.42 0.03 -0.24
NS 3.65 pm (n = 16)
49 2 2 49 f 3 17 f la
47 + 1 48 * 2
32 f 3” 64 + 3’
32 f 1’ 68 z!z2’=
15 f 1
16 -t 1’
15 f lo
5 + za 4 + 1” 3 f oa
6 + 1’ 3 f Ob 2 f Ob
5zkl
420
0 + oa
1 + 0”
321
6 f 1”
2.9 + 0.2
9.0 * 1.2”.b
Results are mean f SE and are in units of counts per 100 except for lamina propria cellularity (%) and collagen table (CT) thickness @ml. “Significantly different from normal subjects by ANOVA (P < 0.05). bSignificantly different from colitis patients with CT < 3.65 pm (P < 0.05).
1794
GASTROENTEROLOGY Vol. 103, No. 6
LEE ET AL.
IO
1600 A 1400 1200-
Discussion
;:ot~ci .
.
.
iooo-
?? ??
800 -
h
-0
5
2
10 CT Thickness
15
20
(p) .
.
800 600 400 -
.
?? ? ?? ? ?? ?? ? ? ? ? ??? ?? ? ?
.
_____._.... .-” ? ______.... .... .
?
200 0-I 40
50 Lamina
60 70 Propria Cellularity
80 (%)
t 90
Figure 3. (A)Correlation of collagen table thickness with stool weight and (B) correlation of lamina propria cellularity with stool weight in patients with microscopic or collagenous colitis. The dashed line is a regression line obtained by the method of least squares using a logarithmic transformation of stool weight.
(Table 2; Figure 3). Lamina propria cellularity and stool weight were significantly correlated, however (I‘= 0.48, P c 0.02; Figure 3). Table 4 shows a comparison of measured collagen table thickness with the pathologist’s global impression of collagen table thickness based on a blinded review of Masson trichrome-stained colon biopsy specimens. Of the 8 patients with normal measured collagen table thickness (~3.65 pm), all were correctly identified as normal by the pathologist. Of the patients with abnormally thick collagen tables, 6 of 16 were correctly identified by the pathologist; these tended to be patients with thicker measured collagen tables (>6 pm), but not all such patients were correctly identified (5 of 11 patients with collagen tables >6 pm were identified as normal). There was agreement between measured collagen table thickness and the pathologist’s assessment in 58% of the cases.
There are several possible explanations for the multimodal frequency distribution of collagen table thickness in patients with microscopic and collagenous colitis. One possibility is that this distribution is attributable to some demographic difference in the populations, such as sex or age.26-26 However, there was no significant difference in sex ratio between patients with normal collagen table thickness and those with abnormally thick collagen tables. Even when only women (the predominant sex in this group of patients) were considered, the frequency distribution was far from normal. Likewise, no effect of age on collagen table thickness was evident, suggesting that collagenous colitis is not just a peculiar response of older individuals to colonic inflammation. Another possibility is that the multimodal distribution of collagen table thickness is caused by differences in the duration of disease. It has been suggested that some patients change with time from a histological picture of microscopic colitis to one of collagenous colitis.‘3*‘Q If this happened very often, there should be a correlation between collagen table thickening and duration of diarrhea; patients with long-standing diarrhea would be more likely to have made this transition. No such correlation was found, suggesting that the metamorphosis from microscopic colitis to collagenous colitis happens infrequently and that collagen table thickening is not merely the nonspecific result of long-standing diarrhea or inflammation. Moreover, because there was no correlation between lamina propria cellularity and collagen table thickness, severity of inflammation per se also does not seem to be responsible for collagen table thickening. The most likely explanation for the multimodal distribution is that the group of colitis patients is made up of two or more distinct populations of patients: one with normal collagen table thickness and one or more with abnormally thick collagen tables. If collagen table thickening was just part of a unitary disease process and microscopic colitis and collage-
Table 4. Comparison of Measured Thickness to Pathologist’s Collagen Table Thickness Trichrome-Stained Colon
Collagen Table Global Impression on Masson Biopsy Specimens
Pathologist’s Measured
thickness
~3.65 pm ~3.65 pm x2 = 2.25; P > 0.05.
Normal 8 10
of
impression Thickened 0 6
December
1992
COLLAGEN TABLE IN MICROSCOPIC/COLLAGENOUS
nous colitis were merely end points on a spectrum of disease with some patients having thicker collagen tables than others, one would expect a unimodal, continuous distribution of collagen table thicknesses.2628 The finding of a multimodal distribution suggests that despite similarities in the type of inflammation present, microscopic colitis and collagenous colitis are distinct conditions, one characterized by no change in collagen table thickness and the other characterized by substantial thickening of the collagen table. The reasons for this distinction may have to do with etiology, pathogenesis, or other differences among the patients. One feature that was not particularly different in patients with and without collagen table thickening was the composition of the inflammatory infiltrate. We could identify no distinguishing features in the type of inflammatory cells between the group of colitis patients with normal average collagen table thickness and those with abnormally thickened collagen tables, except for slightly fewer lymphocytes and macrophages in the patients with abnormally thickened collagen tables. In addition, there was a weak positive correlation between collagen table thickening and numbers of plasma cells, but no correlation with the intensity of inflammation (as reflected by lamina propria cellularity) or the counts of other cell types (eosinophils, fibroblasts, or neutrophils). However, the similarity of inflammatory infiltrates is not strong evidence that microscopic colitis and collagenous colitis are variants of the same disease; the potential patterns of inflammation in the colon are limited and might be shared by several conditions. Our studies also cast some light on the pathogenesis of diarrhea in microscopic colitis and collagenous colitis. Lindstrom postulated that collagen table thickening physically blocked absorption by the colon and thus caused diarrhea.g If this were the case, one might expect that collagen table thickening would correlate with stool weight. We found no such correlation. Instead, there was a significant correlation between lamina propria cellularity and stool weight, suggesting that inflammation and not collagen table thickening per se is responsible for diarrhea in these patients. Our studies also suggest that for research purposes, the distinction between normal and thickened collagen tables needs to be made by measurement rather than just by the pathologist’s subjective assessment. In the present group of patients, biopsy specimens from 1 of 8 patients with normal measured collagen table thicknesses were interpreted as representing collagenous colitis and specimens from 8 of 16 patients with abnormally thick collagen tables were interpreted as microscopic colitis. Even when
COLITIS
1795
trichrome stains were used to highlight the subepithelial collagen table, agreement between measured collagen table thickness and the pathologist’s global assessment was reached only 58% of the time (Table 4). Although this finding might be attributable to the spotty nature of collagen table thickening in some cases or to nosological biases in others, it points out the hazards of subjective pathological analysis in classifying patients for research purposes. However, it is impressive that all but one of the patients identified as having microscopic or collagenous colitis (patient 5, Table 1)had abnormal morphometric analyses; this finding reinforces the notion that it is possible for pathologists to differentiate these conditions from normal. In the past, pathologists have used ocular micrometers to measure the thickness of the subepithelial collagen table in the colon. Reported thicknesses in collagenous colitis have averaged approximately 20 pm and ranged up to 60 pm.2”-24 In the current study, we used a computer-assisted method to measure the average thickness of the collagen table throughout the colon. Our results showed an average collagen table thickness of 9.0 pm among patients with abnormal collagen table thicknesses. Although the difference between our results and those in the literature might reflect different patient populations or the severity of collagen table thickening necessary to prompt a clinical report, we believe it might also reflect an important methodological difference. With the ocular micrometer method the pathologist might select thicker parts of the collagen table for measurement. With the computer-assisted method, the place of measurement is selected by the computer at regular intervals and so selection bias is minimized. Thus, the computer-assisted method is better suited to yield an average value for collagen table thickness throughout the colon. Because collagen table thickening is often a spotty process, the new method may give more representative quantitative results for overall thickening of the collagen table; of course at present such measurement is necessary only for research purposes. References Sylwestrowicz T, Kelly JK, Hwang WS, Shaffer EA. Collagenous colitis and microscopic colitis: the watery diarrhea-colitis syndrome. Am J Gastroenterol 1989;84:783-768, 2. Yardley JH, Lazenby AJ, Giardiello FM, Bayless TM. Collagenous, “microscopic,” lymphocytic, and other gentler and more subtle forms of colitis. Hum Path01 1990;21:1089-1091. 3. Read NW, Krejs GJ, Read MG, Santa Ana CA, Morawski SG, Fordtran JS. Chronic diarrhea ofunknownorigin. Gastroenterology 1980;78:264-271. 4. Kingham JGC, Levison DA, Ball JA, Dawson AM. Microscopic colitis-a cause of chronic watery diarrhoea. Br Med J 1982;285:1601-1604. 5. Bo-Linn GW, Vendrell DD, Lee E, Fordtran JS. An evaluation 1.
1796
GASTROENTEROLOGY
LEE ET AL.
of the significance of microscopic colitis in patients with chronic diarrhea. J Clin Invest 1985;75:1559-1569. 6. Lee E, Schiller LR, Fordtran JS. Quantitation of colonic lamina propria cells by means of a morphometric point-counting method. Gastroenterology 1988;94:409-418. 7. Lazenby AJ, Yardley JH, Giardiello FM, Jessurun J, and Bayless TM. Lymphocytic (“microscopic”) colitis: a comparative histopathologic study with particular reference to collagenous colitis. Hum Path01 1989;20:18-28. 8. Giardiello FM, Lazenby AJ, Bayless TM, Levine EJ, Bias WB, Ladenson PW, Hutcheon DF, Derevjanik NL, Yardley JH. Lymphocytic (microscopic) colitis: clinicopathologic study of 18 patients and comparison to collagenous colitis. Dig Dis Sci
19.
20.
21.
22.
1989;34:1730-1738. 9. Lindstrom
CG. “Collagenous colitis” with watery diarrhea. A new entity. Path01 Eur 1976;11:87-89. 10. Colina F, Solis-Herruzo JA, Munoz-Yague MT, Vazquez G, Perez-Barrios A. Collagenous colitis: the clinical and morphological features. Postgrad Med J 1982;58:390-395. 11. Bogomoletz WV. Collagenous colitis: a clinicopathological review. Surv Dig Dis 1983;1:19-25. 12. Rask-Madsen J, Grove 0, Hansen MGJ, Bukhave K, HenrikNielsen R. Colonic transport of water and electrolytes in a patient with secretory diarrhea due to collagenous colitis. Dig Dis Sci 1983;28:1141-1146.
23. 24.
25. 26. 27.
13. Teglbjaerg PS, Thaysen EH, Jensen HH. Development
enous colitis in sequential ogy 1984;87:703-709.
biopsy specimens.
of collagGastroenterol-
14. Fausa 0, Foerster
histological,
A, Hovig T. Collagenous colitis: a clinical, and ultrastructural study. Stand J Gastroenterol
1985;2O(Suppl 107):8-23.
28. 29.
Vol. 103, No. 6
Jessurun J, Yardley JH, Lee EL, Vendrell DD, Schiller LR, Fordtran JS. Microscopic and collagenous colitis: different names for the same condition? (letter]. Gastroenterology 1986;91:1583-1584. Coverlizza S, Ferrari A, Scevola F, Gemme C, Cavallero M, Spandre M, Risio M, Rossini FP. Clinico-pathological features of collagenous colitis: case report and literature review. Am J Gastroenterol 1986;81:1098-1103. Giardiello FM, Bayless TM, Jessurun J, Hamilton SR, Yardley JH. Collagenous colitis: physiologic and histopathologic studies in seven patients. Ann Intern Med 1987;106:46-49. Jessurun J, Yardley JH, Giardiello FM, Hamilton SR, Bayless TM. Chronic colitis with thickening of the subepithelial collagen layer (collagenous colitis): histopathologic findings in 15 patients. Hum Path01 1987;18:839-848. Rams H, Rogers AI, Ghandur-Mriaymneh L. Collagenous colitis. Ann Intern Med 1987;106:108-113. Lazenby AJ, Yardley JH, Giardiello FM, Bayless TM. Pitfalls in the diagnosis of collagenous colitis: experience with 75 cases from a registry of collagenous colitis at the Johns Hopkins Hospital. Hum Path01 1990;21:905-910. Stampfl DA, Friedman LS. Collagenous colitis pathophysiologic considerations. Dig Dis Sci 1991;36:705-711. Zar JH. Biostatistical analysis, 2nd ed. Englewood Cliffs, NJ: Prentice Hall, 1984. Rimm AA, Hartz AJ, Kalbfleisch JH, Anderson AJ, Hoffmann RG. Basic biostatistics in medicine and epidemiology. New York: Appleton-Century-Crofts, 1980. Sokal RR, Rohlf FJ. Biometry. San Francisco: Freeman, 1981. O’Brien PC, Shampo MA. Statistical considerations for performing multiple tests in a single experiment. 2. Comparisons among several therapies. Mayo Clin Proc 1988;63:816-820.
15. Loo FD, Wood CM, Soergel KH, Komorowski
RA, Cheung H, Gay S, Gay RE. Abnormal collagen deposition and ion transport in collagenous colitis (abstr). Gastroenterology 1985; 88:1481.
16. Kingham JGC, Levison DA, Morson BC, Dawson AM. Collage-
nous colitis. Gut 1986;27:570-577. 17. Palmer KR, Berry H, Wheeler
PJ, Williams CB, Fairclough P, Morson BC, Silk DBA. Collagenous colitis-a relapsing and remitting disease. Gut 1986;27:578-580.
18. Salt WB II, Llaneza PP. Collagenous
colitis: a cause of chronic diarrhea diagnosed only by biopsy of normal appearing coionic mucosa. Gastrointest. Endoscopy 1986;32:421-423,
Received November 27,199l. Accepted June 30,1992. Address requests for reprints to: Lawrence R. Schiller, M.D., Department of Internal Medicine, Baylor University Medical Center, 3500 Gaston Avenue, Dallas, Texas 75246. Supported by U.S. Public Health Service grant 5-ROl-DK3717205 from the National Institute of Diabetes and Digestive and Kidney Diseases and by the Southwest Digestive Disease Foundation. Presented in abstract form at a meeting of the American Gastroenterological Association, May 1989 (Gastroenterology 1989; 96:A293). The authors thank Sharon Michael and Mary Hux for their help in preparing the manuscript.
1992;103:1790-1796
Subepithelial Collagen Table Thickness in Colon Specimens From Patients With Microscopic Colitis and Collagenous Colitis EDWARD LEE, LAWRENCE R. SCHILLER, DORIS VENDRELL, CAROL A. SANTA ANA, and JOHN S. FORDTRAN Departments of Internal Medicine and Pathology, Baylor University Medical Center; Department of Pathology, Veterans Administration Medical Center; and University of Texas Southwestern Medical Center, Dallas, Texas
Microscopic colitis and collagenous colitis are similar conditions that are differentiated by the presence or absence of subepithelial collagen table thickening. To better understand the relationship between these two disorders and the role of collagen table thickening in the pathogenesis of diarrhea, colonic mucosal biopsy specimens from 24 patients with microscopic or collagenous colitis and 9 control subjects were analyzed using a computerassisted morphometric method to evaluate the average thickness of the subepithelial collagen table. The collagen table thickness in colitis patients taken together formed a multimodal rather than a unimodal distribution. There was no tendency for collagen table thickening to increase with age or with duration of symptoms. In general, the types and distribution of inflammatory cells were similar in patients with normal and thickened collagen tables. Stool weight correlated with lamina propria cellularity but not with collagen table thickening. The multimodal distribution of collagen table thickening and the lack of correlation with age, duration of symptoms, or inflammation suggest that microscopic colitis and collagenous, colitis are disalthough the inflammatory crete conditions, changes in the two conditions are similar. Moreover, because stool weight correlates with lamina propria cellularity but not with collagen table thickening, diarrhea probably is caused by the inflammatory changes and not by collagen table thickening per se.
I
n the past 15 years, two new forms of inflammatory bowel disease, microscopic colitis and collagenous colitis, have been described.‘-25 These conditions have many features in common. Both present with watery diarrhea caused by reduced colonic fluid absorption. In both conditions the colon appears normal by barium enema and colonoscopy, but biopsies of the colon show that the lamina propria
contains an inflammatory infiltrate of plasma cells and neutrophils and the epithelium is invaded by lymphocytes and occasional neutrophils. The two conditions are separated histologically only by the presence or absence of subepithelial collagen table thickening. If the subepithelial collagen table is considered abnormally thick, the disorder is labeled collagenous colitis; if not, the disorder is labeled microscopic (or lymphocytic) colitis. However, it is not always possible to determine by routine microscopy whether the collagen layer is thickened, and even expert pathologists may disagree as to whether collagenous colitis or microscopic colitis is present in a given case.” With so many similarities, it has been suggested that these two conditions are variants of the same process with a spectrum of collagen table thickening, perhaps related to age, duration of illness, or severity of inflammation.lg Whether these histological patterns represent different responses to the same disease or two discrete disease processes, the pathophysiological importance of the thickened collagen table is uncertain; some believe that it forms a barrier that inhibits fluid absorption by the colon,’ while others contend that mucosal inflammation is the chief determinant of colonic fluid malabsorption.6 To gain insight into these issues, we carefully measured subepithelial collagen table thickness in the colon and rectum of patients with diagnoses of microscopic or collagenous colitis. Measurements were performed using a novel computer-assisted morphometric method that minimizes observer bias and allows the average collagen table thickness for each patient to be calculated. All biopsy specimens were also evaluated blindly by a previously described computer-assisted morphometric analysis to enumerate the severity of inflammation and the types of inflammatory cells in the lamina propria.” 0 1992
by the American Gastroenterological 0016-5085/92/$3.00
Association
COLLAGEN TABLE IN MICROSCOPIC/COLLAGENOUS COLITIS
December 19%
With knowledge of the average collagen table thickness in each patient, we were then able to address five issues concerning these disorders. First, we plotted collagen table thickness as a frequency distribution, which enabled us to determine whether collagen table thickness in these patients had a unimodal or multimodal distribution.2628 A unimodal, normal distribution would be consistent with the hypothesis of a single disease process; a multimodal distribution would be consistent with either separate disease processes or an extraneous influence that might influence collagen table thickness, such as age, duration of disease, or degree of inflammation. Second, we plotted collagen table thickness as a function of age; a positive correlation would suggest that collagen table thickening is related to aging. Third, we plotted collagen table thickness as a function of duration of illness: a positive correlation would suggest that collagen table thickening is a secondary response to the primary disease rather than an inherent characteristic. Fourth, we plotted collagen table thickness as a function of the severity of mucosal inflammation; a positive correlation would suggest that collagen thickness is determined in part by the degree to which the lamina propria is inflamed. And fifth, we plotted collagen table thickness as a function of daily stool weight; a positive correlation would suggest that collagen table thickening contributes to colonic fluid malabsorption. Materials and Methods Patients and Subjects Demographic and clinical information on 24 patients with clinical pathological diagnoses of microscopic colitis or collagenous colitis is shown in Table 1. Each of these patients had chronic diarrhea (2-348 months’ duration) with stool weights ranging from 246 to 1438 g/24 h. In each case the colonic mucosa was normal when inspected with the colonoscope, but mucosal biopsies were interpreted as indicating microscopic or collagenous colitis. A control group was composed of four normal volunteers who answered classified advertisements and five patients undergoing colonoscopy for evaluation of rectal bleeding or for follow-up after polypectomy. This group included five men and four women aged 37-70 years. Subjects gave written informed consent before participating in this study, which had been approved by the Institutional Review Board for Human Protection. In each subject, colonoscopy was negative and biopsy results were interpreted as normal. Some results from this group have been reported previously.6 Biopsy
Technique
and Processing
Participants were prepared for colonoscopy using Golytely lavage (Braintree Laboratories, Inc., Braintree, MA). When possible, biopsy specimens were obtained, us-
1791
ing Olympus forceps (Olympus Corporation of America, North Chicago, IL), from six regions of the colon: cecum, ascending colon, transverse colon, descending colon, sigmoid colon, and rectum. In some individuals, biopsy specimens were not obtained from all regions. Specimens were fixed in Carson’s buffered formalin, embedded in paraffin, and cut into 6-pm sections on a Leitz 1512 microtome (Australian Biomedical Corp., Melbourne). Sections were stained with H&E or Masson’s trichrome stain. Slides were reviewed by one pathologist (D.V.) who rendered a clinical pathological diagnosis of either microscopic colitis or collagenous colitis and were then coded in such a way that the second pathologist (E.L.), who conducted the morphometric analysis, would be unaware of the identity of the patient or subject, from which group they came (i.e., normal, microscopic colitis, or collagenous colitis), and from which part of the colon the specimens were obtained.
Morphometric
Analysis
A computer-assisted method was developed to measure the thickness of the subepitheiial collagen table without bias to the thickest parts. The &X-magnified image of a trichrome-stained biopsy specimen was conveyed through a Zeiss photomicroscope (Carl Zeiss, Inc., Thornwood, NY) equipped with a Zeiss CCTV downtube adapter to a Hitachi KP-ClOOAcolor video camera (Hitachi Denshi, Ltd., Tokyo, Japan) and was displayed on a CPD1290 RGB color video monitor (Sony, Fujisaua, Japan). A Videometric 150 image analyzer (American Sunovisior, San Diego, CA) was calibrated to the aox image using a stage micrometer and was used to measure the thickness of the subepithelial collagen table. Measurements of collagen table thickness were made at aoym intervals across the surface of well-oriented colon biopsy specimens, so that even if there was an unconscious bias to the initial placement of the specimen for measurement the remaining measurements would be made in an unbiased fashion. A total of 20 measurements were made per region of the colon. In some patients well-oriented sections were not available from all regions. Collagen table thicknesses measured in each region of the colon were averaged together for purposes of statistical analysis. Quantitative morphometric analysis of biopsy specimens to assess inflammation was carried out using a computer-assisted point-counting technique previously validated and described in detail.6 In brief, a dot from a video screen was projected over the magnified image of a welloriented biopsy specimen, and the structure on which the dot landed was identified. The dot was then advanced in a rectilinear pattern across the specimen until 98 points were identified. Other regions of the same biopsy specimen were then evaluated, and the process was repeated until seven areas were counted per specimen from each of the six regions of the colon. More than 4000 points were identified per patient. In some patients, well-oriented sections were not available from all regions of the colon. The results were expressed as the percentage of dots landing on a given structure; this reflects the relative area encompassed by that structure.
1792
GASTROENTEROLOGY
LEE ET AL.
ofMorphometric
Table I. Selected Results
Patient
A@ (Yrl
SEX
1
57
F
2
61
M
Duration of diarrhea WI
Stool weight (S/24 hl
2
1078
30
509
3
50
F
3
564
4
72
F
60
437
5
77
M
2
493
6
55
F
12
317
7
56
F
7
1028
8
58
F
11
1269
9
43
F
84
316
348
429
3
246
10
61
F
11
68
M
12
40
F
192
349
13
69
F
156
1105
14
50
F
156
431
15
42
F
18
424
16
41
F
54
451
17
52
F
5
516
18
70
F
8
600
19
65
F
11
408
20
51
F
240
661
21
44
F
30
573
22
55
F
3
899
23
84
F
54
1438
24
59
F
11
401
Analysis
Clinical pathological diagnosis”
Confidence intervals for control group
MC MC cc MC MC MC MC MC MC MC cc MC MC cc MC MC MC cc cc cc cc cc cc MC
Vol. 103, No. 6
in Individual Patients With Microscopic or Collagenous Colitis Lamina propria cellularity
Neutropbils
Plasma cells
Lymphocytes
MaCUDphases
Fibroblasts
Eosinophils
C&Se,, table thickness
(couilts per 100)
WI
Cm1
7d
15
14b
3
3
6
2.41
7ob
16
13b
3
3
6
2.47
59b
14
6b
3
3
2.68
67b
16
2
7b
4
3
2.76
9
0
3
2
5
3.10
58b
12
2
2
3
6
3.16
68b
18b
5
4
3
2
3.35
68b
16
4
4
3
5
3.54
Ssb
18
4
2
2
5
4.44b
48
61b
11
9b
4
3
5
4.71b
50
12
2
3
3
5
5.07b
72b
19b
5
2
3
3
5.24b
72b
19b
lob
3
2
5
5.88b
66b
15
7b
4
2
4
6.44b
75h
17b
13b
2
2
6
8.66b
63*
16
4
3
3
4
6.98b
74b
17b
10b
3
2
6
7.05b
65b
13
4
4
3
4
7.06b
6gb
19b
3
3
2
3
9.48b
65b
9
0
3
3
7
12.97b
83b
15
2
3
2
3
13.03b
71b
16
3
2
3
4
13.51b
82b
22b
7b
2
2
17.75b
68b
23b
5
3
2
4
37-57
8-16
o-5
O-4
o-4
3-7
18.50b
o-7
1.89-3.65
“MC. microscopic colitis; CC, collaganous colitis. bAbnormal as defined by confidence intervals in control group.
Statistics Data were analyzed by x2 analysis, analysis of variance (ANOVA), group t test, and Pearson’s coefficient of correlation using a computer program (StatPak Gold; Walonick Associates, Inc., Minneapolis, MN).2~2QProbability values of ~0.05 were considered statistically significant. Confidence intervals were calculated as means + ZSD.
Results Collagen table thickness varied little throughout the colon in normal subjects. Collagen table thickness in the cecum, ascending colon, transverse colon, descending colon, sigmoid colon, and rectum averaged 2.83, 2.67, 2.73, 2.75, 2.53, and 3.46 pm, respectively. The average thickness in the rectum was slightly greater than elsewhere in the colon by ANOVA (P < 0.05). For purposes of defining a normal range, all measurements in the normal subjects were the upper limit of normal averaged together; (mean + 2SD) for collagen table thickness was 3.65 pm. Results of measurement of collagen table thickness and morphometric analysis for each patient are shown in Table 1. Patients are arranged in order of average collagen table thickness. For comparison, 95% confidence intervals derived from the control group are also shown. Collagen table thickness was abnormal in 16 of 24 patients; all but 1 of these were women. Of the 8 remaining patients with normal collagen table thickness, 2 were men and 6 were
women. This difference in sex ratio was not statistically significant by x2 analysis. As shown in Figure 1, a frequency polygon of collagen table thickness showed a multimodal distribution, with some colitis patients in the normal range (~3.65 pm) and others with collagen table thicknesses greater than normal. The Kolmogorov-Smirnov statistic for normality was 1.24, strongly suggesting that collagen table thickness in the colitis 701
-fi % 30 r3 B
I I\
Normal Subjects
Colitis Patients
0
2.0
4.0
6.0
8.0 10.0 12.0 CT Thickness @I)
14.0
16.0
18.0
Figure 1. Frequency polygons for percentage of normal subjects or colitis patients with different collagen table thicknesses. The normal subjects form a unimodal population, whereas the colitis patients are a mutimodal population.
December
1992
COLLAGEN
patients was not normally distributed (P < 0.01). This statistic was almost as large (1.19) when only biopsy results from women were considered (P < 0.01). As shown in Figure 2 and Table 2, there was no significant correlation of collagen table thickness with age, duration of diarrhea, or lamina propria cellularity (an index of the overall degree of inflammation). Unadjusted P values for correlation coefficients suggested that collagen table thickness correlated positively with the number of plasma cells and inversely with the numbers of lymphocytes and macrophages; however, these correlations were not statistically significant when Bonferroni adjustment for multiple comparisons was applied.” There was no significant correlation with numbers of eosinophils, fibroblasts, or neutrophils (Table 2). Comparison of morphometric analysis in colitis patients with normal collagen table thickness (~3.65
.
.
. .
lo% .
5-
0’
??
.
40
“30
50
;
. .
-..
60 Age (yr)
.
80
1793
Table 2. Correlations Of Collagen Table Thickness With Clinical Data And Morphometric Results r Clinical data Age Duration of diarrhea Stool weight Morphometric results Lamina propria cellularity Plasma cells Lymphocytes Macrophages Eosinophils Fibroblasts Neutrophils
Unadjusted
0.09 0.02 0.16
NS NS NS
0.33 0.42 -0.42 -0.42 0.03 -0.24
NS 3.65 pm (n = 16)
49 2 2 49 f 3 17 f la
47 + 1 48 * 2
32 f 3” 64 + 3’
32 f 1’ 68 z!z2’=
15 f 1
16 -t 1’
15 f lo
5 + za 4 + 1” 3 f oa
6 + 1’ 3 f Ob 2 f Ob
5zkl
420
0 + oa
1 + 0”
321
6 f 1”
2.9 + 0.2
9.0 * 1.2”.b
Results are mean f SE and are in units of counts per 100 except for lamina propria cellularity (%) and collagen table (CT) thickness @ml. “Significantly different from normal subjects by ANOVA (P < 0.05). bSignificantly different from colitis patients with CT < 3.65 pm (P < 0.05).
1794
GASTROENTEROLOGY Vol. 103, No. 6
LEE ET AL.
IO
1600 A 1400 1200-
Discussion
;:ot~ci .
.
.
iooo-
?? ??
800 -
h
-0
5
2
10 CT Thickness
15
20
(p) .
.
800 600 400 -
.
?? ? ?? ? ?? ?? ? ? ? ? ??? ?? ? ?
.
_____._.... .-” ? ______.... .... .
?
200 0-I 40
50 Lamina
60 70 Propria Cellularity
80 (%)
t 90
Figure 3. (A)Correlation of collagen table thickness with stool weight and (B) correlation of lamina propria cellularity with stool weight in patients with microscopic or collagenous colitis. The dashed line is a regression line obtained by the method of least squares using a logarithmic transformation of stool weight.
(Table 2; Figure 3). Lamina propria cellularity and stool weight were significantly correlated, however (I‘= 0.48, P c 0.02; Figure 3). Table 4 shows a comparison of measured collagen table thickness with the pathologist’s global impression of collagen table thickness based on a blinded review of Masson trichrome-stained colon biopsy specimens. Of the 8 patients with normal measured collagen table thickness (~3.65 pm), all were correctly identified as normal by the pathologist. Of the patients with abnormally thick collagen tables, 6 of 16 were correctly identified by the pathologist; these tended to be patients with thicker measured collagen tables (>6 pm), but not all such patients were correctly identified (5 of 11 patients with collagen tables >6 pm were identified as normal). There was agreement between measured collagen table thickness and the pathologist’s assessment in 58% of the cases.
There are several possible explanations for the multimodal frequency distribution of collagen table thickness in patients with microscopic and collagenous colitis. One possibility is that this distribution is attributable to some demographic difference in the populations, such as sex or age.26-26 However, there was no significant difference in sex ratio between patients with normal collagen table thickness and those with abnormally thick collagen tables. Even when only women (the predominant sex in this group of patients) were considered, the frequency distribution was far from normal. Likewise, no effect of age on collagen table thickness was evident, suggesting that collagenous colitis is not just a peculiar response of older individuals to colonic inflammation. Another possibility is that the multimodal distribution of collagen table thickness is caused by differences in the duration of disease. It has been suggested that some patients change with time from a histological picture of microscopic colitis to one of collagenous colitis.‘3*‘Q If this happened very often, there should be a correlation between collagen table thickening and duration of diarrhea; patients with long-standing diarrhea would be more likely to have made this transition. No such correlation was found, suggesting that the metamorphosis from microscopic colitis to collagenous colitis happens infrequently and that collagen table thickening is not merely the nonspecific result of long-standing diarrhea or inflammation. Moreover, because there was no correlation between lamina propria cellularity and collagen table thickness, severity of inflammation per se also does not seem to be responsible for collagen table thickening. The most likely explanation for the multimodal distribution is that the group of colitis patients is made up of two or more distinct populations of patients: one with normal collagen table thickness and one or more with abnormally thick collagen tables. If collagen table thickening was just part of a unitary disease process and microscopic colitis and collage-
Table 4. Comparison of Measured Thickness to Pathologist’s Collagen Table Thickness Trichrome-Stained Colon
Collagen Table Global Impression on Masson Biopsy Specimens
Pathologist’s Measured
thickness
~3.65 pm ~3.65 pm x2 = 2.25; P > 0.05.
Normal 8 10
of
impression Thickened 0 6
December
1992
COLLAGEN TABLE IN MICROSCOPIC/COLLAGENOUS
nous colitis were merely end points on a spectrum of disease with some patients having thicker collagen tables than others, one would expect a unimodal, continuous distribution of collagen table thicknesses.2628 The finding of a multimodal distribution suggests that despite similarities in the type of inflammation present, microscopic colitis and collagenous colitis are distinct conditions, one characterized by no change in collagen table thickness and the other characterized by substantial thickening of the collagen table. The reasons for this distinction may have to do with etiology, pathogenesis, or other differences among the patients. One feature that was not particularly different in patients with and without collagen table thickening was the composition of the inflammatory infiltrate. We could identify no distinguishing features in the type of inflammatory cells between the group of colitis patients with normal average collagen table thickness and those with abnormally thickened collagen tables, except for slightly fewer lymphocytes and macrophages in the patients with abnormally thickened collagen tables. In addition, there was a weak positive correlation between collagen table thickening and numbers of plasma cells, but no correlation with the intensity of inflammation (as reflected by lamina propria cellularity) or the counts of other cell types (eosinophils, fibroblasts, or neutrophils). However, the similarity of inflammatory infiltrates is not strong evidence that microscopic colitis and collagenous colitis are variants of the same disease; the potential patterns of inflammation in the colon are limited and might be shared by several conditions. Our studies also cast some light on the pathogenesis of diarrhea in microscopic colitis and collagenous colitis. Lindstrom postulated that collagen table thickening physically blocked absorption by the colon and thus caused diarrhea.g If this were the case, one might expect that collagen table thickening would correlate with stool weight. We found no such correlation. Instead, there was a significant correlation between lamina propria cellularity and stool weight, suggesting that inflammation and not collagen table thickening per se is responsible for diarrhea in these patients. Our studies also suggest that for research purposes, the distinction between normal and thickened collagen tables needs to be made by measurement rather than just by the pathologist’s subjective assessment. In the present group of patients, biopsy specimens from 1 of 8 patients with normal measured collagen table thicknesses were interpreted as representing collagenous colitis and specimens from 8 of 16 patients with abnormally thick collagen tables were interpreted as microscopic colitis. Even when
COLITIS
1795
trichrome stains were used to highlight the subepithelial collagen table, agreement between measured collagen table thickness and the pathologist’s global assessment was reached only 58% of the time (Table 4). Although this finding might be attributable to the spotty nature of collagen table thickening in some cases or to nosological biases in others, it points out the hazards of subjective pathological analysis in classifying patients for research purposes. However, it is impressive that all but one of the patients identified as having microscopic or collagenous colitis (patient 5, Table 1)had abnormal morphometric analyses; this finding reinforces the notion that it is possible for pathologists to differentiate these conditions from normal. In the past, pathologists have used ocular micrometers to measure the thickness of the subepithelial collagen table in the colon. Reported thicknesses in collagenous colitis have averaged approximately 20 pm and ranged up to 60 pm.2”-24 In the current study, we used a computer-assisted method to measure the average thickness of the collagen table throughout the colon. Our results showed an average collagen table thickness of 9.0 pm among patients with abnormal collagen table thicknesses. Although the difference between our results and those in the literature might reflect different patient populations or the severity of collagen table thickening necessary to prompt a clinical report, we believe it might also reflect an important methodological difference. With the ocular micrometer method the pathologist might select thicker parts of the collagen table for measurement. With the computer-assisted method, the place of measurement is selected by the computer at regular intervals and so selection bias is minimized. Thus, the computer-assisted method is better suited to yield an average value for collagen table thickness throughout the colon. Because collagen table thickening is often a spotty process, the new method may give more representative quantitative results for overall thickening of the collagen table; of course at present such measurement is necessary only for research purposes. References Sylwestrowicz T, Kelly JK, Hwang WS, Shaffer EA. Collagenous colitis and microscopic colitis: the watery diarrhea-colitis syndrome. Am J Gastroenterol 1989;84:783-768, 2. Yardley JH, Lazenby AJ, Giardiello FM, Bayless TM. Collagenous, “microscopic,” lymphocytic, and other gentler and more subtle forms of colitis. Hum Path01 1990;21:1089-1091. 3. Read NW, Krejs GJ, Read MG, Santa Ana CA, Morawski SG, Fordtran JS. Chronic diarrhea ofunknownorigin. Gastroenterology 1980;78:264-271. 4. Kingham JGC, Levison DA, Ball JA, Dawson AM. Microscopic colitis-a cause of chronic watery diarrhoea. Br Med J 1982;285:1601-1604. 5. Bo-Linn GW, Vendrell DD, Lee E, Fordtran JS. An evaluation 1.
1796
GASTROENTEROLOGY
LEE ET AL.
of the significance of microscopic colitis in patients with chronic diarrhea. J Clin Invest 1985;75:1559-1569. 6. Lee E, Schiller LR, Fordtran JS. Quantitation of colonic lamina propria cells by means of a morphometric point-counting method. Gastroenterology 1988;94:409-418. 7. Lazenby AJ, Yardley JH, Giardiello FM, Jessurun J, and Bayless TM. Lymphocytic (“microscopic”) colitis: a comparative histopathologic study with particular reference to collagenous colitis. Hum Path01 1989;20:18-28. 8. Giardiello FM, Lazenby AJ, Bayless TM, Levine EJ, Bias WB, Ladenson PW, Hutcheon DF, Derevjanik NL, Yardley JH. Lymphocytic (microscopic) colitis: clinicopathologic study of 18 patients and comparison to collagenous colitis. Dig Dis Sci
19.
20.
21.
22.
1989;34:1730-1738. 9. Lindstrom
CG. “Collagenous colitis” with watery diarrhea. A new entity. Path01 Eur 1976;11:87-89. 10. Colina F, Solis-Herruzo JA, Munoz-Yague MT, Vazquez G, Perez-Barrios A. Collagenous colitis: the clinical and morphological features. Postgrad Med J 1982;58:390-395. 11. Bogomoletz WV. Collagenous colitis: a clinicopathological review. Surv Dig Dis 1983;1:19-25. 12. Rask-Madsen J, Grove 0, Hansen MGJ, Bukhave K, HenrikNielsen R. Colonic transport of water and electrolytes in a patient with secretory diarrhea due to collagenous colitis. Dig Dis Sci 1983;28:1141-1146.
23. 24.
25. 26. 27.
13. Teglbjaerg PS, Thaysen EH, Jensen HH. Development
enous colitis in sequential ogy 1984;87:703-709.
biopsy specimens.
of collagGastroenterol-
14. Fausa 0, Foerster
histological,
A, Hovig T. Collagenous colitis: a clinical, and ultrastructural study. Stand J Gastroenterol
1985;2O(Suppl 107):8-23.
28. 29.
Vol. 103, No. 6
Jessurun J, Yardley JH, Lee EL, Vendrell DD, Schiller LR, Fordtran JS. Microscopic and collagenous colitis: different names for the same condition? (letter]. Gastroenterology 1986;91:1583-1584. Coverlizza S, Ferrari A, Scevola F, Gemme C, Cavallero M, Spandre M, Risio M, Rossini FP. Clinico-pathological features of collagenous colitis: case report and literature review. Am J Gastroenterol 1986;81:1098-1103. Giardiello FM, Bayless TM, Jessurun J, Hamilton SR, Yardley JH. Collagenous colitis: physiologic and histopathologic studies in seven patients. Ann Intern Med 1987;106:46-49. Jessurun J, Yardley JH, Giardiello FM, Hamilton SR, Bayless TM. Chronic colitis with thickening of the subepithelial collagen layer (collagenous colitis): histopathologic findings in 15 patients. Hum Path01 1987;18:839-848. Rams H, Rogers AI, Ghandur-Mriaymneh L. Collagenous colitis. Ann Intern Med 1987;106:108-113. Lazenby AJ, Yardley JH, Giardiello FM, Bayless TM. Pitfalls in the diagnosis of collagenous colitis: experience with 75 cases from a registry of collagenous colitis at the Johns Hopkins Hospital. Hum Path01 1990;21:905-910. Stampfl DA, Friedman LS. Collagenous colitis pathophysiologic considerations. Dig Dis Sci 1991;36:705-711. Zar JH. Biostatistical analysis, 2nd ed. Englewood Cliffs, NJ: Prentice Hall, 1984. Rimm AA, Hartz AJ, Kalbfleisch JH, Anderson AJ, Hoffmann RG. Basic biostatistics in medicine and epidemiology. New York: Appleton-Century-Crofts, 1980. Sokal RR, Rohlf FJ. Biometry. San Francisco: Freeman, 1981. O’Brien PC, Shampo MA. Statistical considerations for performing multiple tests in a single experiment. 2. Comparisons among several therapies. Mayo Clin Proc 1988;63:816-820.
15. Loo FD, Wood CM, Soergel KH, Komorowski
RA, Cheung H, Gay S, Gay RE. Abnormal collagen deposition and ion transport in collagenous colitis (abstr). Gastroenterology 1985; 88:1481.
16. Kingham JGC, Levison DA, Morson BC, Dawson AM. Collage-
nous colitis. Gut 1986;27:570-577. 17. Palmer KR, Berry H, Wheeler
PJ, Williams CB, Fairclough P, Morson BC, Silk DBA. Collagenous colitis-a relapsing and remitting disease. Gut 1986;27:578-580.
18. Salt WB II, Llaneza PP. Collagenous
colitis: a cause of chronic diarrhea diagnosed only by biopsy of normal appearing coionic mucosa. Gastrointest. Endoscopy 1986;32:421-423,
Received November 27,199l. Accepted June 30,1992. Address requests for reprints to: Lawrence R. Schiller, M.D., Department of Internal Medicine, Baylor University Medical Center, 3500 Gaston Avenue, Dallas, Texas 75246. Supported by U.S. Public Health Service grant 5-ROl-DK3717205 from the National Institute of Diabetes and Digestive and Kidney Diseases and by the Southwest Digestive Disease Foundation. Presented in abstract form at a meeting of the American Gastroenterological Association, May 1989 (Gastroenterology 1989; 96:A293). The authors thank Sharon Michael and Mary Hux for their help in preparing the manuscript.
