Quantitation of lymphocytes in endomyocardial biopsies: Use and limitations of antibodies to leukocyte common antigen

Quantitation of Lymphocytes in Endomyocardial Biopsies: Use and Limitations of Antibodies to Leukocyte Common Antigen STUART J. SCHNITi-, MD,* PETER S. CIANO, MD,* AND FREDERICK J. SCHOEN, MD, PHDt cyte necrosis is also present, 6 the accurate diagnosis of myocarditis largely d e p e n d s on distinguishing mononuclear inflammatory cells from other cellnlar components of the myocardial interstitium. This distinction is of clinical importance since some patients, when diagnosed with myocarditis, are treated with immunosuppressive therapy, which itself may be associated with serious complications. 6 Recently, antibodies to a variety of leukocyte antigens, including antibodies to specific lymphocyte subsets, have been used to identify and quantitate interstitial mononuclear cells in frozen sections of end o m y o c a r d i a l biopsy samples. TM H o w e v e r , this method requires a separate frozen biopsy specimen, provides suboptimal morphologic preservation, and is not suitable for retrospective analyses. A technique to identify mononuclear inflammatory cells reliably in routinely prepared (i.e., formalin-fixed, paraffine m b e d d e d ) endomyocardial biopsy samples would therefore be highly desirable. Before such a method is adopted, however, its specificity and reproducibility must be evaluated. To address these issues, we investigated the use of a commercially available antibody to the pan-leukocyte marker leukocyte common antigen (LCA) to identify mononuclear inflammatory cells in the myocardial interstitium using an immunoperoxidase technique on routinely processed endomyocardial biopsy specimens. The goals of this study were 1) to determine the interobserver reproducibility in the identification and quantitation of interstitial inflammatory cells using LCA-immunoperoxidase-stained sections and to compare the results to those obtained on hematoxylin-eosin-stained sections, and 2) to determine the general usefulness of antibodies to LCA in the routine evaluation of endomyocardial biopsy samples.

The difficulty in distinguishing lymphocytes from other cells in the myocardial interstitium on hematoxylin-eosin-stained sections has contributed to the variable incidence of myocarditis diagnosed by endomyocardial biopsy. We investigated the value of immunoperoxidase staining using antibodies to leukocyte common antigen (LCA) for identifying interstitial mononuclear ceils in formalin-fixed, paraffin-embedded, endomyocardial biopsy samples of the right ventricle obtained by bioptome from 50 consecutive autopsies. In 42 cases with nonspecific findings or no disease, two observers independently quantitated interstitial lymphocytes on hematoxylin-eosin-stained sections and LCA-positive cells on immunoperoxidase sections. Mean values for lymphocytes on hematoxylin-eosin-stained sections were significantly different for the two observers (5.8/mm 2 vs. 8.9/ m m ~, P < 0.001). In contrast, mean values for LCA-positive cells on immunoperoxidase sections were similar (13.8/mm ~ vs. 15.1/ m m 2, P = not significan0, lnterobserver concordance was excellent for LCA-immunoperoxidase sections (r = 0.83), exceeding that obtained for hematoxylin-eosin-stained sections (r = 0.63). Intraobserver correlations between lymphocytes identified on hematoxylin-eosin-stained sections and LCA-positive cells on immunoperoxidase sections were poor (r = 0.28 and 0.14). LCAimmunoreactive cells included, in addition to lymphocytes, mast cells and histiocytes. We conclude that there is considerable interobserver discordance in identifying lymphocytes on hematoxylin-eosin-stained sections of endomyocardial biopsy samples. While LCA-immunoperoxidase staining reduces interobserver variability, results must be interpreted with caution since this antibody stains other leukocytes in addition to lymphocytes. HUM PATHOL 18:796-800, 1987.

Variable criteria for histologic diagnosis have contributed to the wide range in the reported frequency of myocarditis among patients who undergo endomyocardial biopsy. However, a feature common to all diagnosticcriteria is the presence of interstitial inflammatory cells, predominantly lymphocytes. 1-5 T h e absolute number of interstitial lymphocytes that distinguishes active myocarditis from nonmyocarditic disease and from normal is uncertain. 6 Although many cardiac pathologists recommend that the diagnosis of active myocarditis be made only when myo-

MATERIALSAND METHODS

From the Departments of Pathology, *Beth Israel Hospital and tBrigham and Women's Hospital, and Harvard Medical School, Boston, Massachusetts.Acceptedfor publication6 October I986. Supported in part by a grant from the Cordis Corporation, Miami, Florida. Address correspondenceand reprint requests to Dr. Schnitt: Department of Pathology,Beth Israel Hospital, 330 BrooklineAvenue, Boston, MA 02215. 0046-8177/87 $0.00 + .25 796

R i g h t v e n t r i c u l a r e n d o m y o c a r d i a l biopsy samples were obtained from 50 consecutive, unrestricted autopsy cases. For each case, five fragments o f e n d o m y o c a r d i u m were obtained from tile right side of the interventricular septum near tile cardiac apex using a disposable bioptome. This sampling technique was chosen to simulate, as closely as possible, the location, specimen size, and configuration of clinical endomyocardial biopsy samples obtained during cardiac catheterization. The biopsy specimens

QUANIffATION OF LYMPHOCYIES IN ENDOMYOCARDL~L BIOPSIES[Schnitt et al.]

for each case were fixed in formalin and embedded in a single paraffin block, and 5-1xm-thick sections were obtained for hematoxylin-eosin staining. Review of these sections revealed acute myocardial infarction in seven cases and leukemic infiltration in one case; these were excluded from further analysis. In the remaining 42 cases, the myocardium was either essentially normal (n = 21) or showed mild, nonspecific changes (i.e., lipofilscin deposits, myocyte hypertrophy, interstitial or endocardial fibrosis) (n = 21). For these cases, consecutive 5-~m-thick sections were stained with antibodies to LCA (Dako Corp., Santa Barbara, California) and the a v i d i n - b i o t i n c o m p l e x i m m u n o p e r o x i d a s e t e c h n i q u e 9 with a methyl green counterstain. Negative controls were obtained by substituting mouse myeloma protein for the primary antibody. Sections o f formalin-fixed, paraffin-embedded tonsil served as positive controls. T h e llematoxylin-eosin- and LCA-stained sections for each case were examined independently by two observers (SJS and PSC) who had no knowledge of other pathologic or clinical data. The LCA-stained sections were coded to prevent comparison with the corresponding hematoxylin-eosin-stained sections. Each observer recorded the number of interstitial cells presumed to be lymphocytes ill 20 consecutive high-power microscopic fields (each with an area of 0.17 mm 9) on the hematoxylin-eosin-stained sections and determined the mean number of these cells per high-power field for each case. Similarly, each observer recorded the number of LCA-positive interstitial cells in 20 high-power fields and determined the mean n u m b e r o f LCA-positive cells per highpower field for each case. LCA-immunoreactive cells in blood vessels or associated with extravasated erythrocytes were excluded (fig. 1). Autopsy records were reviewed for each case to determine the interval between the time of death and time of autopsy (postmortem interval), patient age,

TABLE t. Quantitative Evaluation of Interstitial Lymphocytes on Hematoxylin-Eosin-stained Sections and Leukocyte Common Antigen-positive Cells on Immunoperoxidase-stained Sections of Endomyocardial Biopsies* LCA-positive L y m p h o c y t e s / m m2 Cells/mm ~ ( H e m a t o x y l i n - E o s i n ) ( I m m u n o p e r o x i d a s e ) P Value Observer I Observer 2 P value

RESULTS

Tile study population consisted of 19 men and 23 women ranging in age from 37 to 98 )'ears (mean, 69.8). The postmortem interval ranged from three to 41 hours (mean, 12). T h e mean heart weight for the 42 cases was 409 g (range, 240 to 700). Routine sections of the heart obtained at autopsy revealed no pathologic changes in 11 cases, nonspecific hypertrophic and degenerative changes in 25 cases, and acute infarcts in areas away from the biopsy sites in six cases.

A broad range of values was obtained for the 42 cases for both the number of presumed lymphocytes oil hematoxylin-eosin-stained sections (1.2 to 24.6/ mm 2) and the number of LCA-positive cells on immunoperoxidase-stained sections (1.8 to 55.9/mm2). T h e mean numbers of lymphocytes detected on hematoxylin-eosin staining in the 42 cases differed significantly for the two observers. In contrast, the mean numbers of LCA-positive cells determined on i m m u n o p e r o x i d a s e - s t a i n e d sections were similar (table 1). Furthermore, interobserver concordance on LCA-stained sections was excellent (r = 0.83), exceeding that obtained on hematoxylin-eosin-stained sections (r = 0.63) (fig. 2). Table 1 also indicates that for both observers, the mean numbers of LCA-positive cells on immunoperoxidase-stained sections were substantially higher than tile mean numbers of presumed lymphocytes detected on hematoxylin-eosin-stained sections. Int r a o b s e r v e r correlations b e t w e e n h e m a t o x y l i n eosin- and immunoperoxidase-stained sections were correspondingly poor (observer 1, r = 0.28; observer 2, r = 0.14). This, in part, relates to the fact that the antibody to LCA employed in this study reacts with other types of mononuclear leukocytes in addition to lymphocytes.ll

4'

B

e e 9

f ,

<0.001 <0.001

sex, heart weight, cause of death, and general and cardiac pathologic findings. Mean values were compared by the Wilcoxon paired sample test or the Mann-Whitney test (for unpaired values). Correlations were calculated by the rank correlation method of Spearman. l~

"tr

.

13.8 _ 1.5 15.1 - 1.3 NS

ABBREVIATIONS: LCA, leukocyte c o m m o n a n t i g e n ; NS, not significant. * M e a n values --- SE for 49 cases.

t

9

5.8 - 0.50 8.9 - 0.66 <0.001

r ~

-

#

FIGURE t. Leukocyte c o m m o n antigen-positive cells in an interstitial blood vessel. Cells such as these were not included in the quantitation. [Immunoperoxidase stain with methyl green counterstain, x 600.]

797

HUMAN PATHOLOGY

30 f

Volume 18, No. 8 [August 1987]

//'/

25

9

/i

60

//# , /

50

s,/

o,I 20 ]. ,--=. 15 t

//

II~

" .

=

"o,;///y

.

~

iI

,,,,,~ 30

./r

.r

//

ii

//

//

e~

", , , "/ " | "=, 9 I i~9 .9 5 [/ , 9 o/ 49 9

O 20

10F

0

t .#.#/ I /

0

A

/pl

I1~

,.Q

o

40

// // P

pl

p//

I

5

9

.

q~..r

10 y,

I

10

I

I

15 20 Observer 1

I

I

25

30

0

(

I

I

0

10

20

S

I

I

30 40 Observer 1

I

i

50

60

FIGURE 2. Correlation plots demonstrating interobserver concordance for the evaluation of lymphocytes counts (per m m 2] on hematoxylin-eosin-stained sections CA, r = 0.63] and leukocyte c o m m o n antigen-positive cell counts [per m m ~} on immunoperoxidase-stained sections [B, r = 0.83].

On the basis of their morphology and staining patterns, LCA-positive cells in the myocardial interstitium could be separated into three groups: lymphocytes, mast cells, and elliptical to spindle-shaped cells (fig. 3). Lymphocytes were round, had scant cytoplasm, and showed distinct peripheral reactMty for LCA. Mast cells were round to ovoid, had small central nuclei, and showed coarse, granular, cytoplasnfic LCA intmunoreactivity. T h e third group consisted of elongated elliptical to spindle-shaped cells with central nuclei and fine, diffuse, cytoplasmic or surface staining. Ahhough the nature of these cells could not be determined with certainty, they most likely represent histiocytes. Lymphocytes constituted the majority of LCA-positive cells in only 10 of the 42 cases (24 per cent). In most cases (64 per cent), the predominant LCA-immunoreactive cell type was elliptical to spindle-shaped. In five cases (12 per cent), mast cells were the most n u m e r o u s LCA-positive cells. It should be noted that mast cells have been previously reported to react directly with avidinbiotin-peroxidase complexes in tissue sections. 12 In fact, in 30 cases (71 per cent), occasional mast cells were seen on the immunoperoxidase-negative control slides in which the primary antibody was replaced by mouse myeloma protein. However, in each case, these appeared to represent only a minority of the mast cells present on the corresponding LCAstained section. T h e relationship between the quantitative evaluations and the pathologic findings on hematoxylineosin-stained sections of the endomyocardial biopsy samples is pre~ented in table 2. T h e r e was no significant difference in the number of presumed lymphocytes per square millimeter on hematoxylin-eosinstained sections among cases in which the biopsy samples showed normal myocardium and those in

which the biopsy samples showed nonspecific chronic changes. However, cases in which chronic changes were present showed a significantly higher number of LCA-positive cells per square millimeter than cases in which the biopsies were normal. Neither the n u m b e r of lymphocytes on hematoxylin-eosin-stained sections nor the n u m b e r of LCA-positive cells on immunoperoxidase-stained sections correlated with patient age, sex, postmortem interval, heart weight, cause of death, or general or cardiac pathologic findings in routine autopsy sections.

DISCUSSION The results of this study document a considerable interobserver variability in the recognition of lymphocytes on hematoxylin-eosin-stained sections o f endomyocardial biopsy samples. Such variability probably has contributed to the lack of agreement on tile incidence of myocarditis in patients undergoing endomyocardial biopsy, n-5 Prior investigators have noted that a variety of cells in the myocardial interstitium (e.g., endothelial cells, pericytes, and fibroblasts) may be mistaken for mononuclear inflammatory cells on hematoxylin-eosin-stained sections. 5,7 However, to our knowledge, ours is the first objective study of the interobserver variation in the recognition of myocardial interstitial lymphocytes. T h e data also demonstrate that interobserver reproducibility in the identification o f interstitial mononuclear cells in endomyocardial biopsy samples can be improved by the use of antibodies to leukocyte common antigen (LCA) and the immunoperoxidase technique. Linder et al. 7 have reported that antibodies to lymphocyte surface markers can be used on

798

QUANTITAT1ONOF LYMPHOCYIESIN ENDOMYOCARDIALBIOPSIES[Schnitt et al.)

!

9

%

~r-

~..

~

\

FIGURE 3. Morphologic groups of leukocyte common antigen [LCA)-Immunoreactive cells in the myocardial interstitium, a, Lymphocytes. Round cells with peripheral immunoreactivity for LCA. b, Mast cell. Note coarse granular cytoplasmic immunoreactivity for LCA and small central nucleus, c, Spindle-shaped ovoid ceil. The nucleus is elliptical, and there is fine diffuse cytoplasmic staining for !_CA.(A//, immunoperoxidase stain with methyl green counterstain, x 600.}

J 9

II

" 4

tt'- ~.. , . ,

$

~,

.

e

.:q

.~.

.~' o

:

0 .? ~

..........

frozen sections to identify and categorize interstitial lymphocytes in the myocardium accurately and to distinguish them from other cell types. However, freezing requires that a separate biopsy specimen be obtained. In addition, morphologic preservation on frozen sections is suboptimal, and it may be extremely difficult to distinguish immunoreactive cells in the interstitium from those in blood vessels. Finally, such a method is not applicable to archival material for retrospective studies9 Therefore, a method to identify mononuclear inflammatory cells reliably in the myocardial interstitium in routinely processed biopsy samples would have great practical value9 T h e reactivity of the antibody used in this study has been well characterized in paraffin-embedded tissue. ~ Consistent immunostaining has been demonstrated in lymphocytes and mast cells, and variable staining has been observed in histiocytes and plasma cells. In addition, the antigen recognized by this antibody can be reliably demonstrated in tissue obtained at autopsy even up to 72 hours postmortem. 1~ This suggests that the results of the present study could be extended to biopsy samples obtained under the usual clinical circumstances. It is important to emphasize that in most of our cases, the predominant LCA-immunoreactive cells 799

-. . . . . .

~

......

-:2

. . . . . . . .

were not lymphocytes. Therefore, caution is required in interpreting LCA:immunoperoxidase-stained sections on routinely processed endomyocardial biopsy samples in which tile diagnosis of myocarditis is snsp e c t e d on h e m a t o x y l i n - e o s i n - s t a i n e d sections. Careful attention must be paid to the size, shape, and pattern of staining of the LCA-positive cells so that immunoreactive cells of other types are not mistaken for lymphocytes. Additional markers that may be useful in paraffin-embedded sections to categorize further interstitial cells include markers for histiocytes (~rantitrypsin, ~x-antichymotrypsin, and lysozyme), which could be employed when the nature of the LCA-positive cells is in doubt. Another factor that has contributed to the confusion surrounding the diagnosis of myocarditis on endomyocardial biopsy samples is the paucity of information regarding the number of lymphocytes and other mononuclear cells present in the myocardial interstitium normally and in nonmyocarditic disease. Edwards et al., 5 using hematoxylin-eosin-stained sections of 170 biopsy samples, considered a mean lymphocyte count of fewer than 5/high-power field not to represent myocarditis. This value is equivalent to a mean lymphocyte count of approximately 25 to 30/mm 2 of myocardium. 7 Linder et al. 7 studied I0

HUMAN PATHOLOGY

Volume 18, No. 8 [August t987]

TABLE 2. Quantitative Evaluation of Interstitial Cells Related to Pathologic Findings in Hematoxylin-Eosinstained Sections of Endomyocardial BiopsySamples*

m o n o n u c l e a r i n f l a m m a t o r y cells and for distinguishing them from other cellular components of the myocardial interstitium.

Pathologic Findings in Biopsies

Normal (n = 21)

Lymphocytes (hematoxylineosin),/mm 2 LCA-positive cells (immunoperoxidase),/mm~

Nonspecific Chronic Changes (n = 21)

P Value

Acknowledgments. The authors thank Ellie Manseau and Jodi Newman for performing the immunoperoxidase stains. The bioptomes used in this study were provided by the Cordis Corporation, Miami, Florida. REFERENCES

6.8

8.0

NS

11.8

17.6

0.01

ABBREVIATIONS: LCA, leukocyte common antigen; NS, not significant. * Mean values calculated from values obtained by both observers for 42 cases.

autopsy hearts and four right ventricular endomyocardial biopsy samples to determine the n u m b e r of lymphocytes in normal myocardium. These investigators, using antibodies to the pan-leukocyte marker T-200 on frozen sections, found an average 8.9 T-200-positive cells/ram 9 at autopsy and an average of 13.1 immunoreactive cells/mm~ in the biopsy specimens. O u r results, obtained with an antibody to LCA (which is analogous to T-200 antigen) tt on routinely processed biopsy specimens (table 1), are similar to those of Linder et al. 7 Thus, taken together, the results of Edwards et al., 5 Linder et al., 7 and the present study provide a frame of reference for the numbers of lymphocytes and LCA-positive cells ttmt may be encountered in cardiac biopsy samples from patients without active myocarditis. Nevertheless, quantitative lymphocyte counts must be interpreted with considerable caution and in the context of other morphologic findings. For example, clinically significant myocarditis may be focal in distribution, 5 and lymphocyte clusters may be present within the myocardium of patients with idiopathic chronic congestive cardiomyopathy. 14 Therefore, this teclmique would be most ttseful as an adjunct in the evahtation of abnormal endomyocardial biopsy samples in which the nature of the interstitial cellsis difficultto determine with certainty on hematoxylin-eosin examination. In s u m m a r y , the i m m u n o h i s t o c h e m i c a l and quantitative m e t h o d described here uses a single commercially available antibody preparation, is applicable to formalin-fixed, paraffin-embedded tissue, and is valuable for retrospective analyses. Despite some limitations, this m e t h o d provides a useful means for objectively and reproducibly identifying

800

1. Waller B, Slack J, Dillon J, et al: Rarity of myocarditis in endomyocardial biopsy in patients with new onset idiopathic congestive heart failure. A histologic and uhrastructural, immunohistochenlical and viral serologic analysis of 72 patients [Abstract]. J Am Coil Cardiol 7:120A, 1986 2. Zee-Cheng C-S, Tsai C, Palmer D, et al: High incidence of myocarditis by endomyocardial biopsy in patients with idiopathic congestive cardioxnyopatlW. J Am Coil Cardiol 3:63, 1984 3. Fenoglio J J, Ursell PC, Kellogg CF, et al: Diagnosis and classification of myocarditis by endomyocardial biopsy. N Engl J Med 308:12, 1983 4. Dec GW, Palacios IF, Fallon JT, et ah Active myocarditis in the spectrtun of acute dilated cardiomyopatlfies. Clinical features, Ifistologic correlates and clinical outcome. N Engl J Med 312:885, 1985 5. Edwards W, Holmes D, Reeder G: Diagnosis of active lymphocytic myocarditis by endomyocardial biopsy. Quantitative criteria for light microscopy. ,Mayo Clin Proc 57:419, 1982 6. Aretz HT, Billinghana ME, Edwards WD, et al: Myocarditis. A histopathologic definition and classification. Am J Cardiovasc Pathol 1:3, 1986 7. l[.inder J, Cassling R, Rogler W, et al: Imnmnolfistochemical characterization of lymphocytes in uninflamed myocardium. Implications for myocarditis. Arch Pathol Lab Med 109:917, 1985 8. Cassling R, Linder J, Sears T, et aI: Quantitative evaluation of inflammation in biopsy specimens from idiopathically failing or irritable hearts: experience in 80 pediatric and aduh patients. Am HeartJ 110:713, 1985 9. ttsu SM, Raine L, Fanger H: The use of avidin-biotin-peroxidase complex (ABC) in immunoperoxidase teclmiques: a comparison between ABC anti unlabeled antibody (PAP) procedures. J Histochem Cytochem 29:577, 1981 10. Zar JH: Biostatistical Analysis. Euglewood Cliffs, New Jersey, Prentice-Hall, Inc., 1974 11. Kurtin PJ, Pinkus GS: Leukocyte common antigen--a diagnostic discriminant between hematopoietic and non-hematopoietic neoplasms in paraffin sections using monoclonal antibodies: correlation with immunologic studies and uhrastructural localization. HUM PATIIOL 16:353, 1985 12. Hsu SM, Raine L: The use of avidin-biotin-peroxidase complex (ABC) in diagnostic and research pathology. In DeLellis R (ed): Advances in hnmunohistocllenfistry. New .~ork, Masson Publislfing, USA, Inc., 1984 13. Pallesen G, Knudsen L: Leukocyte antigens in hunmn postmortem tissues: their preservation and loss as demonstrated by monoclonal antibody immtmolfistological staining. Histopathology 9:791, 1985 14. Tazelaar HD, Billingham ME: Leukocytic infiltrates in idiopathic dilated cardiomyopathy: a source of confusion with active myocarditis. Am J Surg Pathol 10:405, 1986