APMIS ZOO: 1 1 15-1 122, 1992
Immunohistology of skin and rectum biopsies in bone marrow transplant recipients RITVA LESKINEN', EERO TASKINEN', LIISA VOLIN, TAPANI RUUTU2 and PEKKA HAYRY' 'Transplantation Laboratory and 'Third Department of Medicine, University of Helsinki, Helsinki, Finland
Leskinen, R.. Taskinen, E., Volin, L., Ruutu, T. & Hayry, P. Immunohistology of skin and rectum biopsies in bone marrow transplant recipients. APMIS 100: 11 15-1122, 1992. We have studied histological and immunohistological specimens of 39 skin biopsies from 21, and 30 rectal biopsies from 17 bone marrow transplant recipients. The biopsies were taken before transplantation, during acute and chronic graft-versus-host disease (GVHD), and at times with no GVHD. In biopsies taken during cutaneous aGVHD grade I to 111, epithelial changes were seen in 16/23 biopsies. The cutaneous infiltrates during aGVHD consisted of CD2-, CD4-, CD8- and FMC-33-positive cells both in the epithelium and in the dermis. CD57-positive NK cells were also detected in most biopsies. During chronic GVHD the cutaneous cellular infiltrates were similar to those seen in moderate aGVHD, i.e. both CD4- and CD8-positive lymphoid cells were present. When the biopsy was taken after the beginning of corticosteroid treatment for aGVHD, or at times when the patient did not have GVHD symptoms, the cellular infiltrates were considerably smaller in the dermis. During clinical intestinal aGVHD mucosal epithelial changes were relatively uncommon; instead, increased numbers of both CD4- and CD8-positive lymphocytes in the lamina propria (LP) were seen in 1 1 / 13 samples. During chronic GVHD the number of CD4-positive cells exceeded that of CDS-positive cells in the LP, and the large lymphoid infiltrates also reached the muscularis mucosae. In rectal biopsies the differences were not so prominent because most of the pretransplant biopsies showed CD2-, CD4-, CD8- and CD57-positive lymphocytes both in the lamina propria and epithelium. Key words: Skin and gut immunohistology; acute graft-versus-host disease. Ritva Leskinen, Transplantation Laboratory, University of Helsinki, Haartmaninkatu 3, SF 00290 Helsinki, Finland.
Different skin symptoms are very common after bone marrow transplantation (BMT). Graftversus-host disease and exanthemas caused by irradiation, medication and viruses are hard to distinguish. The clinical diagnosis of early GVHD can be difficult and the histological changes of aGVHD in the skin are not specific. Epidermal basal cell degeneration with spongiosis and bulla formation is present in more severe forms, and lymphoid cell infiltration into the dermoepidermal junction has been considered typical for aGVHD (8). Epidermal ab-
Received July 10, 1992. Accepted October 8, 1992.
normalities alone, without involvement of the dermis, are frequently seen in post-transplant biopsies of patients with no aGVHD symptoms (3). In addition to aGVHD, gut problems can be caused by several other factors such as infections and drugs, and it is difficult to distinguish clinically the different causes of diarrhoea. The most severe injury caused by aGVHD seen at autopsy is often located in the terminal ileum, but because of the difficulties in obtaining ileal biopsies, the findings of intestinal GVHD are based on biopsies of rectum, where the injuries are not so pronounced. In the gut, apoptotic lesions in the crypts, glandular dilatation or mucosal denudation and 1115
LESKINEN et a/.
lymphoid cell infiltration in the lamina propria (LP) are seen histologically during aGVHD (8). The aim of the present study was to characterize using immunohistological methods the types of infiltrating mononuclear cells in the skin and gut in GVHD, bearing in mind the possible use of these findings in the differential diagnosis of GVHD.
MATERIALS AND METHODS Pat ien ts All patients were transplanted for a malignant haematological illness: 1 1 had chronic myeloid leukaemia, nine acute myeloid leukaemia, five acute lymphatic leukaemia, and one Burkitt’s lymphoma. They were conditioned with high-dose cyclophosphamide (120 mg/kg) and total body irradiation (10-12 Gy). All but two patients received fractionated irradiation. Three patients were given methotrexate and 19 patients cyclosporin A as GVHD prophylaxis. Biopsies Thirty-nine 3 mm punch skin biopsies were taken from 21 BMT recipients: four biopsies before BMT, 23 during clinical skin aGVHD, (14 without corticosteroid treatment, indicated as aGVHD, S-, and nine with corticosteroid treatment, aGVHD, S +), seven during chronic GVHD (cGVHD), and five from recipients without clinical GVHD symptoms at the time of the biopsy. Thirty rectum biopsies were taken from 17 BMT recipients: 10 before BMT, 14 during aGVHD (five without and nine with corticosteroid treatment), two during cGVHD, and four from recipients without clinical GVHD. Twelve of the patients in the aGVHD group had skin and gut symptoms, and two had skin and liver GVHD. Histopathological technique Formalin-fixed biopsy specimens were embedded in paraffin and 14 pm-thick sections were stained with haematoxylin and eosin. Histopalhological grading Skin biopsies were histologically graded according to Lerner et al. (8): Grade I: vacuolar degeneration of epidermal basal cells and acantocytes, Grade 11: vacuolar changes with spongiosis and dyskeratosis, Grade 111: epidermolysis and bulla formation; Grade IV: total denudation of the epidermis. In the gut the histopathological grading was as follows: Grade I: focal dilatation and degeneration of the mucosal glands; Grade 11: focal or diffuse loss of intestinal glands, Grade 111: mucosal denudation. If the grade I11 type change is diffuse, the classification is Grade 1V (8).
1116
Processing of the biopsies f o r immunohistochemistry The biopsies were immersed in Tissue-Tek (LabTek Products, Division of Miles Laboratories, Inc., Naperville, Ill.) and snap-frozen in liquid nitrogen. Frozen sections were cut at 4-5 pm and labelled by indirect immunoperoxidase staining; air-dried acetone-fixed sections were also fixed with chloroform for 30 min, incubated with a monoclonal mouse antihuman antibody for 30 min, and washed twice in TRIS-NaCI buffer, pH 7.4. As a second antibody, a peroxidase-conjugated rabbit anti-mouse antibody was used. After two washings with TRIS-NaC1 buffer, the slides were incubated with peroxidase-conjugated goat anti-rabbit antibody. The slides were exposed to a mixture of AEC (3-amino-9-ethyl carbazole) and H202 in 0.05 M acetate buffer, pH 5.0, counterstained for one min in Mayer’s haemalum and mounted with Aquamount. The following monoclonal antibodies were used: CD2: OKTl1 (Ortho), CD4: T4 (Coulter), CD8: OKT8 (Ortho), CD20: B1 (Coulter), and CD57: Leu7 (Becton Dickinson). The anti-monocyte monoclonal antibodies FMC 17 and FMC 33 were gifts from Dr Heddy Zola, University of Adelaide, South Australia. The number of infiltrating cells in the immunohistological stainings was graded from - to 4 + , No or only occasional positive cells were graded -; slightly increased numbers of positive cells were graded 1 . Moderate infiltrates were coded as 2 + , large infiltrates as 3 +, and very intense infiltrates as 4 + .
+
RESULTS Skin biopsies The immunohistology of skin biopsies is summarized in Tables 1 and 2. Biopsies taken before transplantation and before conditioning ( n = 4 ) The histology of all biopsies was normal. In immunoperoxidase stainings, only occasional CD2 , CD4+ and CD8 + lymphocytes could be detected in the dermis; these were in close contact with capillaries. In one patient the CD8+ cells were focally increased in the upper dermis (grade 1+). Stainings with CD57 and CD20 antibodies were negative.
+
Biopsies taken during aGVHD according to clinical judgement ( n =23) Fourteen of these biopsies were taken when the patients did not have corticosteroid treatment for aGVHD. Nine biopsies were taken during corticosteroid treatment. The epidermal
CUTANEOUS AND RECTAL IMMUNOHISTOLOGY IN GVHD
TABLE 1. Immunohistological findings in skin biopsies of BMT patients Patient groups Pre aGVH,S - aGVH,S cGVH 717 414 3/14 3I9 0 0 3 2 I 0 0 3 0 3 I1 0 0 5 1 I11
+
Biopsy histology
Epidermis CD2+ cells
CD4+ cells
CD8+ cells
CD57+ cells
Dermis CD2+ cells
CD4+ cells
CDS+ cells
+ ++ +++ + ++ +++ + ++ +++ + ++ +++ -
+ ++ +++ ++++ -
+ ++ +++ ++++ + ++ +++ ++++
414 0 0 0 414 0 0 0 4/4 0 0 0 313 0 0 0
2/12 5 4 1 2/12 3 7 0 2/12 4 6 0 9/12 2 1 0
414 0 0 0 0 414 0 0 0 0 314
0/12 5 3 3 1 01 12 2 7 3 0 2112 3 2 5 0
1
0 0 0
noGVH 315 1 1 0
I 15 3
315 2
1 0 1 /6 1 4 0
0 0 I/5 2 2
116 4 1 0 315 2 0 0
315 2 0 0 515 0 0 0
318
015
4 0
2 3 0 0 016 2 4 0 0
315 2 0 0 0 315 2 0 0 0 215 2
5/8
2 1 0 418 3 0 1 518 2 1 0 5/6 1 0 0
1
0 018
5 2 1
0 218 4 1 1 0
016
3 3 0 0
0
1 0 0
5/6 013 6/12 215 315 1 2 3 3 0 0 3 0 0 0 0 0 0 0 0 0 0 0 0 0 pre = before transplantation, aGVH,S - =during acute GVHD without corticosteroid treatment, aGVH,S + = during acute GVHD with corticosteroid treatment, cGVH = during chronic GVHD, noGVH =during quiescent period without GVHD. Biopsy histology indicates the histopathological findings in the epithelium graded as: 0 = no changes, I =vacuolar degeneration of basal cells, I1 =vacuolar changes with spongiosis and dyskeratosis, I11 = epidermolysis and bulla formation. The number of positive cells in immunohistochemical stainings is expressed as follows: - =not detectable or only occasional, + = slightly increased numbers, =moderate infiltrates, + = intense infiltrates, and + =very large infiltrates. CD57+ cells
++ +
+ ++ +++ ++++
++
++
1117
LESKINEN el al.
histology was normal in 6/23 sections. Other biopsies displayed grade 1-111 changes. Immunohistological stainings were done from 20 biopsies for CD2, CD4 and CD8 antibodies. In the biopsies taken in the absence of corticosteroid treatment, epidermal CD2 + , CD4 + and CD8 lymphocytes were seen in 10/ 12 biopsies. The number of CD 57-positive cells was increased in the epidermis in three biopsies. In the dermis, CD2+ and CD4+ infiltrates with intensity grades of 1 to + 4 were seen in all biopsies, and CD8 infiltrates in 10/ 12 biopsies. In all but two biopsies the number of all these types of lymphocytes was increased compared to biopsies taken before BMT. CD57positive cells were seen in the dermis in 6/12 biopsies. Corticosteroid treatment effectively influenced the number of both epidermal and dermal lymphocytes; most infiltrates were grade I + in severity. The results of detailed examination of the biopsies taken from cutaneous lesions during aGVHD, which displayed epidermal changes typical for aGVHD and where immunohistological stainings for most subclasses were available, are given in Table 2. The lymphoid infiltrates were more intense in the dermis than in the epidermis, but in most biopsies the same subclasses of mononuclear cells were seen in both
+ +
Day Post BMT +I2 +25 +49
+57 $34 +22 $77 +28 +28 +I9 +13 +10 +24 +53
Histology I I1 I I11 I 111
I1 I1 111
I1
I I1
I1 I11
structures. In the epidermis, FMC 33-positive cells were seen of grade 1 to 2+ infiltrates, but in the dermis they formed large infiltrates.
+
Biopsies taken during cG VHD ( n = 7) The histology was normal in all cases, except for parakeratosis noticed in one sample. In spite of this, CD4+ or CD8 T cells were found in all biopsies both in the epidermis and dermis. The epidermal and dermal clusters of cells consisted mostly of CD4+ lymphocytes, but the number of CD2+, CD8+ and CD57f lymphocytes was also increased in most biopsies when compared with pretransplant biopsies. The dermal infiltrates were graded + 1 to + 2 in intensity.
+
Biopsies taken during periods of no clinical GVHD (n=5) Three biopsies were histologically normal, one was classified as grade I GVHD (taken at autopsy) and one as grade I1 GVHD (the patient had suffered from a mild skin GVHD three weeks before). CD4+ lymphocytes ( + 1 to + 2 in intensity) were seen in the epidermis in 415 samples, CD8 + and CD2 + cells ( + 1 in intensity) in only two samples. In the dermis, CD2 , CD4+ and CD8+ cells were seen in slightly increased numbers only around the capillaries.
+
TABLE 2. Biopsies from skin lesions during aGVHD Epidermis Dermis CD2 CD4 CD8 CD57 FMC33 CD2 CD4 CD8
-
+ ++ ++ + + ++ ++ +++ ++- +
-
+ ++ ++ ++ ++
-
++ ++ ++ ++ + + ++ + + + + ++ + + +- ++++ ++ -
+
-
+ + ++ + + ++ + + ++ + -
-
+
-
-
+
+ ++++ +++ +++ + ++ +++ +- +
+
+++ ++ ++ ++ +++ +++ ++ ++ ++ ++ + +++ +++ + ++
-
+++ +++ +++ + +++ +++ ++ ++ ++ + + +++ +
CD57 FMC33 CS
-
+ ++ ++ ++ -
-
+-
++ +++ ++ ++ +++ ++
-
-
+ + +++ + +
Histology indicates epidermal histopathological changes: I: vacuolar degeneration of basal cells, 11: vacuolar changes together with spongiosis and dyskeratosis, 111: epidermolysis and bulla formation. The numbers of positive cells are expressed with symbols: - =not detectable or only occasional, + =slightly increased numbers, + + =moderate infiltrates, =large infiltrates, + + + + =very intense infiltrates. CS = corticosteroid treatment at the time of biopsy.
+++
1118
CUTANEOUS AND RECTAL IMMUNOHISTOLOGY IN GVHD
TABLE 3. Immunohistological findings in rectal biopsies of BMT patients Biopsy histology Patient groups Pre aGVH,S aGVH,S cGVH noGVH 0 61 10 315 419 1I2 214 1 4 2 3 0 2 11 0 0 0 1 0 111 0 0 1 0 0 IV 0 0 1 0 0 Epithelium CD2+ cells 318 1 I5 217 012 1 I4 5 2 4 1 3 0 2 1 1 0 0 0 0 0 0 CD4+ cells 619 115 217 012 214 2 4 4 2 2 1 0 0 0 0 0 0 0 1 0 CD8+ cells 1I 9 015 017 012 1/3 6 3 3 0 2 2 2 3 1 0 1 1 0 0 0 CD57+ cells 419 214 1 i6 oi2 0/4 5 2 4 2 4 0 0 1 0 0 0 0 0 0 0 Lamina propria CD2+ cells 018 014 018 012 014 6 2 2 0 3 5 0 1 2 2 0 0 I 1 0 0 0 0 1 0 CD4+ cells 019 015 018 012 014 5 I 1 0 2 6 0 2 3 4 1 0 0 1 0 0 0 0 2 0 CD8+ cells 119 015 012 013 018 7 2 3 0 1 4 0 2 1 2 0 1 1 2 0 0 0 0 0 0 CD57+ cells 719 013 215 012 1/ 4 2 3 3 2 3 0 0 0 0 0 0 0 0 0 0 ++++ 0 0 0 0 0 Biopsy histology indicates the histopathological findings in gut epithelium. 0 =no changes, I = focal dilatation and degeneration of mucosal glands, I1 = focal or diffuse loss of mucosal glands, 111= mucosal denudation. The patient groups and the coding of the intensity of cellular infiltrates are the same as in Table I .
+
+ ++ +++ + +++ ++
+ ++
+++
+ ++ +++
+ ++ +++ ++++
+ ++ +++ ++++
+ ++ +++ ++++
+ ++ +++
1119
LESKINEN
CD20+ B cells were not found in any of the skin biopsies. Rectal biopsies The results of rectal biopsies are summarized in Tables 3 and 4 Pretransplant biopsies ( n = 10) The histology was mostly normal, but slight dilatation of the crypts was seen in four biopsies. CD2 +, CD8 + and CD57 + lymphocytes were seen in the epithelium in most biopsies. No or only occasional CD4+ cells were seen in the epithelium in 6/9 biopsies. In the lamina propria CD2 +, CD4 + and CD8 + cells were detected in practically all biopsies stained with these antibodies. In one sample CD4+ cells were increased to grade 3 + and in three samples to 2 + infiltrates, while CD8+ cells of grade 2 + intensity were seen in one sample. These patients had been given cytostatic drugs only some days or weeks before the biopsy. Biopsies taken during aG VHD ( n = 14) Slight dilatation of crypts, coded as I, was seen in five biopsies. In one biopsy epithelial denudation, grade 111, and in one biopsy grade IV was seen. In spite of the slight epithelial changes, the total number of lymphocytes in the lamina propria was increased in 11/ 14 biopsies. In the glandular epithelium there was no clear difference in the numbers of CD2-, CD8- and CD57-positive lymphocytes as compared to the pretransplant biopsies; instead, CD4 + lymphocytes were seen in small infiltrates in 9/ 12 biopsies. In the lamina propria the intensity of CD4+ cells increased up to 2 + in 1 1 / 13 biopsies, and the differences in the intensities of CD2+ and CD8+ infiltrates were even more clear compared to pretransplant biopsies. The corticosteroid treatment did not influence the changes in rectal biopsies as markedly as in skin biopsies. Biopsies taken during chronic GVHD ( n = 2) In the epithelium the number of CD8+ cells was increased. In the lamina propria the number of CD4 + cells exceeded that of CD8 cells and the infiltrates were more intensive than during aGVHD, up to grades + 3 and +4. In these two biopsies the strong lymphocytic infiltrate also reached the muscularis mucosae layer.
+
1120
et a1
Biopsies taken during no clinical G VHD ( n = 4 ) These findings did not differ from the pretransplant biopsies. No CD20+ B cells were detected in the rectal biopsies. The detailed results from rectal biopsies taken during acute GVHD are summarized in Table 4.
DISCUSSION We studied skin and rectum biopsies taken from BMT patients before transplantation, during GVHD, and during quiescent periods. The cellular infiltrates both in cutaneous and intestinal structures were seen considerably more clearly in immunohistological than in the usual histological stainings. The immunohistopathological changes are often local and therefore we did not count the cell numbers per area, but estimated the cell numbers in the whole biopsy and graded the findings from - to 4 + . This study demonstrates that typical features of the maculopapular skin lesions of aGVHD are infiltrates of both CD4+ and CD8+ lymphocytes and FMC33+ monocytes both in the dermis and the epidermis. The changes in rectal biopsies are more difficult to interpret because also the pretransplant biopsies contained a considerable number of both CD4+ and CD8 + cells. Moreover, most of the patients were receiving corticosteroid treatment at the time of the rectum biopsy, which could have diminished the intensity of cellular infiltrates. During intestinal GVHD the increase in the CD2 + and CD8 + infiltrates in the lamina propria compared to the pretransplant biopsies was the most typical finding. No marked differences in the epithelial infiltrates of CD8 + or CD57 + cells were seen during GVHD compared to the pre-BMT biopsies; instead, the numbers of CD4 + cells were slightly increased. Our results resemble those of Elliot et al. who found an increase in CD2 + , CD4 + and CD8 + lymphocytes in skin biopsies exhibiting aGVHD changes (2), but are in disagreement with some earlier studies where the lymphoid cells in cutaneous GVHD are reported to be predominantly CD8+ (4, 5, 6 , 7, 10). In some previous studies the antibody used for CD4+ cells was
CUTANEOUS AND RECTAL IMMUNOHISTOLOGY IN GVHD
Day post BMT 69 56 89 50 20 28 40 53 27 28
Histology
+ + + + ++ + ++ +
111
0 0 0 I 0 I I 0 0 0 IV I
+118
+ 123 + 110
TABLE 4. Rectal biousies taken durina aG V H D Epithelium Lamina propria CD2 CD4 CD8 CD57 CD2 CD4 -
-
+ + ++ ++ ++ ++ + + * +
+
-
++ ++ + ++ -
+ * +
+
+ + +
++ ++
-
++ ++ + ++ ++ ++ ++ * +++
+
++ -
+ + +
*
+
+ ++ + ++ ++ ++ ++ + ++ + +++ ++ ++ + ++ + + ++ ++ ++ +++ +++ ++ ++
CD8
CD57
CS
++ ++ + ++ ++ +++ ++ + + +++ ++ +++
+ +
+ + + + + +
-
-
++ + + +
-
-
-n -
- if
Histology indicates epithelial histopathological changes coded as: 0 =no changes, I = focal dilatation and degeneration of mucosal glands, I1 =focal or diffuse loss of intestinal glands, 111= mucosal denudation, IV = diffuse mucosal denudation. The number of positive cells is graded as in Table 2 CS = corticosteroid treatment * =epithelial denudation these patients had skin and liver aGVHD, the others clinical skin and gut aGVHD. I'
OKT4 (Ortho) ( 5 , 6, 7), whereas we have used T4 (Coulter) because of the better antibody function in the immunoperoxidase staining. The staining methods in these earlier studies were also different, the immunofluorescence method usually being employed (2, 6, 7). The increase in the number of monocytes and NK cells in skin GVHD in our study is in agreement with Elliot et al. (2). There are only a few reports of the immunohistology of rectal biopsies during aGVHD (1, 11). Dilly & Sloane reported an increase in the number of CD8+ lymphocytes in the lamina propria with no change in CD4 + cells ( I ) . Weisdorf et al. found increased numbers of CD8 , as well as CD57+ lymphocytes, both in the epithelium and lamina propria during GVHD (11). We could not observe any increase in CD57 lymphoid cells during GVHD either in the gut epithelium or in the lamina propria. Most of the CD57+ cells seen in our biopsies in every patient group were obviously the neuroendocrine cells of the gut epithelium. Otherwise our results are concordant with the two earlier reports, and perhaps the different prophylaxis for GVHD in these studies could have influenced the results. The immunohistology of skin biopsies might be useful in the differential diagnosis of skin
+
+
problems after BMT because it is difficult to see the lymphoid cell infiltrates in normal histology stainings. On the other hand, only epidermal histological changes are frequently seen after BMT. Total body irradiation of 1000 rad causes very mild and transient epithelial atypia, which vanishes within a week. Cyclophosphamide and total body irradiation together can produce dyskeratosis with or without satellitosis, but these changes disappear during the three weeks after transplantation. The epithelial damage caused by chemotherapy and irradiation cannot be distinguished from the epithelial changes caused by aGVHD (9). Therefore the diagnosis of aGVHD should also include dermal lymphoid cell infiltration (3, 9). The cellular composition of the infiltrate in the dermis in aGVHD is typical and immunohistological stainings can help in the diagnosis of atypical skin lesions of aGVHD. The rectal biopsy histology in intestinal aGVHD is unspecific and the changes are often local. The pathogenesis of intestinal symptoms during aGVHD is also more complicated than in the skin. The initial intestinal lesion caused by effector T cells can facilitate the invasion of Gram-negative bacteria into the lamina propria, and endotoxin from these bacteria could enhance the immunological reaction by stimulat-
1121
LESKINEN er (11.
ing the release of interleukin-1, interferon and tumour necrosis factor. Thus the intensity and quality of the cellular infiltrates of the rectal mucosa may not have as clear a correlation with the clinical GVHD symptoms as in the skin. In immunohistological stainings the increase in CD8+ lymphocytes in the lamina propria during aGVHD is a constant though not always prominent finding. In patients whose intestinal symptoms are severe enough to need treatment and where the GVHD diagnosis is uncertain, rectal biopsy with immunohistological stainings may prove to be of some help, although information on the cellular infiltrates in infectious gut problems, which constitute the main differential diagnosis, is so far very limited.
REFERENCES Dilly, S. A . & Sloane, J . I?: Changes in rectal leucocytes after allogeneic bone marrow transplantation. Clin. Exp. Immunol. 67: 151-158, 1987. Elliot, C. J., Sloune, J. II, Pallett, G. D.& Sundemon, K. V: Cutaneous leucocyte composition after human allogeneic bone marrow transplantation: relationship to marrow purging, histology and clinical rash. Histopathology 12: 1-16, 1988. Elliot, C. J., Sloane, J. I?, Sanderson, K. V , Vincent, M . , Shepherd, V & Powles, R.: The histological diagnosis of cutaneous graft versus host disease: relationship of skin changes to marrow purging and other clinical variables. Histopathology 11: 145-155, 1987. Guyotat. D., Mauduit, G., Chouvet, B., Kanitakis,
I122
5.
6.
7.
8.
9.
10.
11.
J., Vu Van, H . , Fiere, D. & Thivolet, V: A sequential study of histological changes in the skin after allogenic bone marrow transplantation. Transplantation 41: 340-342, 1986. Janossy, G., Montano, L., Selby, W S., Duke, O., Panayi, G., Lampert, I., Thomas, J. A . . Granger, S., Bofll, M., Tidman, N . , Thomas, N. C. & Goldstein, G.: T cell abnormalities in tissue lesions developing during autoimmune disorders, viral infection and graft-versus-host disease. J. Clin. Immunol. (Suppl.); 2: 42-56, 1982. Kuye, V N., Neumann, I? M., Kersey, J., Goltz, R. W , Baldridge, B. D.. Michael, A . E & Platt. J. L.: Identity of immune cells in graft-versushost disease of the skin. Am. J. Pathol. 116: 436440, 1984. Lampert, I. A , , Janossy, G., Suitters, A . J., Janossy, G., Thomas, J. A., Palmer, S. & Gordon Smith, E.; Immunological analysis of the skin in graft versus host disease. Clin. Exp. Immunol. SO: 123-131, 1982. Lerner, K. G., Kao, G. E , Storb, R., Buckner, C. D., Clift, R. A . & Thomas, E. D.: Histopathology of graft-vs.-host reaction (GvHR) in human recipients of marrow from HL-A-matched sibling donors. Transplant. Proc. 6: 367-371, 1974. Sale, G. E., Lerner, K. G.. Barker, E. A , , Shulmun, H. M . & Thomas, E. D.: The skin biopsy in the diagnosis of acute graft-versus-host disease in man. Am. J. Pathol. 89: 621-634, 1977. Sloane, J. I?, Thomas, J. A., Imrie, S. F , Easton. D. E & Powless, R. L.: Morphological and immunohistological changes in the skin in allogeneic bone marrow recipients. J. Clin. Pathol. 37: 919-930, 1984. Weisclorf; S. A.. Roy, J., Snover, D.,Platt, J. L. & Weisdor- D.J.: Inflammatory cells in graftversus-host disease on the rectum: immunopathologic analysis. Bone Marrow Transplant. 7: 297-301, 1991.
Immunohistology of skin and rectum biopsies in bone marrow transplant recipients RITVA LESKINEN', EERO TASKINEN', LIISA VOLIN, TAPANI RUUTU2 and PEKKA HAYRY' 'Transplantation Laboratory and 'Third Department of Medicine, University of Helsinki, Helsinki, Finland
Leskinen, R.. Taskinen, E., Volin, L., Ruutu, T. & Hayry, P. Immunohistology of skin and rectum biopsies in bone marrow transplant recipients. APMIS 100: 11 15-1122, 1992. We have studied histological and immunohistological specimens of 39 skin biopsies from 21, and 30 rectal biopsies from 17 bone marrow transplant recipients. The biopsies were taken before transplantation, during acute and chronic graft-versus-host disease (GVHD), and at times with no GVHD. In biopsies taken during cutaneous aGVHD grade I to 111, epithelial changes were seen in 16/23 biopsies. The cutaneous infiltrates during aGVHD consisted of CD2-, CD4-, CD8- and FMC-33-positive cells both in the epithelium and in the dermis. CD57-positive NK cells were also detected in most biopsies. During chronic GVHD the cutaneous cellular infiltrates were similar to those seen in moderate aGVHD, i.e. both CD4- and CD8-positive lymphoid cells were present. When the biopsy was taken after the beginning of corticosteroid treatment for aGVHD, or at times when the patient did not have GVHD symptoms, the cellular infiltrates were considerably smaller in the dermis. During clinical intestinal aGVHD mucosal epithelial changes were relatively uncommon; instead, increased numbers of both CD4- and CD8-positive lymphocytes in the lamina propria (LP) were seen in 1 1 / 13 samples. During chronic GVHD the number of CD4-positive cells exceeded that of CDS-positive cells in the LP, and the large lymphoid infiltrates also reached the muscularis mucosae. In rectal biopsies the differences were not so prominent because most of the pretransplant biopsies showed CD2-, CD4-, CD8- and CD57-positive lymphocytes both in the lamina propria and epithelium. Key words: Skin and gut immunohistology; acute graft-versus-host disease. Ritva Leskinen, Transplantation Laboratory, University of Helsinki, Haartmaninkatu 3, SF 00290 Helsinki, Finland.
Different skin symptoms are very common after bone marrow transplantation (BMT). Graftversus-host disease and exanthemas caused by irradiation, medication and viruses are hard to distinguish. The clinical diagnosis of early GVHD can be difficult and the histological changes of aGVHD in the skin are not specific. Epidermal basal cell degeneration with spongiosis and bulla formation is present in more severe forms, and lymphoid cell infiltration into the dermoepidermal junction has been considered typical for aGVHD (8). Epidermal ab-
Received July 10, 1992. Accepted October 8, 1992.
normalities alone, without involvement of the dermis, are frequently seen in post-transplant biopsies of patients with no aGVHD symptoms (3). In addition to aGVHD, gut problems can be caused by several other factors such as infections and drugs, and it is difficult to distinguish clinically the different causes of diarrhoea. The most severe injury caused by aGVHD seen at autopsy is often located in the terminal ileum, but because of the difficulties in obtaining ileal biopsies, the findings of intestinal GVHD are based on biopsies of rectum, where the injuries are not so pronounced. In the gut, apoptotic lesions in the crypts, glandular dilatation or mucosal denudation and 1115
LESKINEN et a/.
lymphoid cell infiltration in the lamina propria (LP) are seen histologically during aGVHD (8). The aim of the present study was to characterize using immunohistological methods the types of infiltrating mononuclear cells in the skin and gut in GVHD, bearing in mind the possible use of these findings in the differential diagnosis of GVHD.
MATERIALS AND METHODS Pat ien ts All patients were transplanted for a malignant haematological illness: 1 1 had chronic myeloid leukaemia, nine acute myeloid leukaemia, five acute lymphatic leukaemia, and one Burkitt’s lymphoma. They were conditioned with high-dose cyclophosphamide (120 mg/kg) and total body irradiation (10-12 Gy). All but two patients received fractionated irradiation. Three patients were given methotrexate and 19 patients cyclosporin A as GVHD prophylaxis. Biopsies Thirty-nine 3 mm punch skin biopsies were taken from 21 BMT recipients: four biopsies before BMT, 23 during clinical skin aGVHD, (14 without corticosteroid treatment, indicated as aGVHD, S-, and nine with corticosteroid treatment, aGVHD, S +), seven during chronic GVHD (cGVHD), and five from recipients without clinical GVHD symptoms at the time of the biopsy. Thirty rectum biopsies were taken from 17 BMT recipients: 10 before BMT, 14 during aGVHD (five without and nine with corticosteroid treatment), two during cGVHD, and four from recipients without clinical GVHD. Twelve of the patients in the aGVHD group had skin and gut symptoms, and two had skin and liver GVHD. Histopathological technique Formalin-fixed biopsy specimens were embedded in paraffin and 14 pm-thick sections were stained with haematoxylin and eosin. Histopalhological grading Skin biopsies were histologically graded according to Lerner et al. (8): Grade I: vacuolar degeneration of epidermal basal cells and acantocytes, Grade 11: vacuolar changes with spongiosis and dyskeratosis, Grade 111: epidermolysis and bulla formation; Grade IV: total denudation of the epidermis. In the gut the histopathological grading was as follows: Grade I: focal dilatation and degeneration of the mucosal glands; Grade 11: focal or diffuse loss of intestinal glands, Grade 111: mucosal denudation. If the grade I11 type change is diffuse, the classification is Grade 1V (8).
1116
Processing of the biopsies f o r immunohistochemistry The biopsies were immersed in Tissue-Tek (LabTek Products, Division of Miles Laboratories, Inc., Naperville, Ill.) and snap-frozen in liquid nitrogen. Frozen sections were cut at 4-5 pm and labelled by indirect immunoperoxidase staining; air-dried acetone-fixed sections were also fixed with chloroform for 30 min, incubated with a monoclonal mouse antihuman antibody for 30 min, and washed twice in TRIS-NaCI buffer, pH 7.4. As a second antibody, a peroxidase-conjugated rabbit anti-mouse antibody was used. After two washings with TRIS-NaC1 buffer, the slides were incubated with peroxidase-conjugated goat anti-rabbit antibody. The slides were exposed to a mixture of AEC (3-amino-9-ethyl carbazole) and H202 in 0.05 M acetate buffer, pH 5.0, counterstained for one min in Mayer’s haemalum and mounted with Aquamount. The following monoclonal antibodies were used: CD2: OKTl1 (Ortho), CD4: T4 (Coulter), CD8: OKT8 (Ortho), CD20: B1 (Coulter), and CD57: Leu7 (Becton Dickinson). The anti-monocyte monoclonal antibodies FMC 17 and FMC 33 were gifts from Dr Heddy Zola, University of Adelaide, South Australia. The number of infiltrating cells in the immunohistological stainings was graded from - to 4 + , No or only occasional positive cells were graded -; slightly increased numbers of positive cells were graded 1 . Moderate infiltrates were coded as 2 + , large infiltrates as 3 +, and very intense infiltrates as 4 + .
+
RESULTS Skin biopsies The immunohistology of skin biopsies is summarized in Tables 1 and 2. Biopsies taken before transplantation and before conditioning ( n = 4 ) The histology of all biopsies was normal. In immunoperoxidase stainings, only occasional CD2 , CD4+ and CD8 + lymphocytes could be detected in the dermis; these were in close contact with capillaries. In one patient the CD8+ cells were focally increased in the upper dermis (grade 1+). Stainings with CD57 and CD20 antibodies were negative.
+
Biopsies taken during aGVHD according to clinical judgement ( n =23) Fourteen of these biopsies were taken when the patients did not have corticosteroid treatment for aGVHD. Nine biopsies were taken during corticosteroid treatment. The epidermal
CUTANEOUS AND RECTAL IMMUNOHISTOLOGY IN GVHD
TABLE 1. Immunohistological findings in skin biopsies of BMT patients Patient groups Pre aGVH,S - aGVH,S cGVH 717 414 3/14 3I9 0 0 3 2 I 0 0 3 0 3 I1 0 0 5 1 I11
+
Biopsy histology
Epidermis CD2+ cells
CD4+ cells
CD8+ cells
CD57+ cells
Dermis CD2+ cells
CD4+ cells
CDS+ cells
+ ++ +++ + ++ +++ + ++ +++ + ++ +++ -
+ ++ +++ ++++ -
+ ++ +++ ++++ + ++ +++ ++++
414 0 0 0 414 0 0 0 4/4 0 0 0 313 0 0 0
2/12 5 4 1 2/12 3 7 0 2/12 4 6 0 9/12 2 1 0
414 0 0 0 0 414 0 0 0 0 314
0/12 5 3 3 1 01 12 2 7 3 0 2112 3 2 5 0
1
0 0 0
noGVH 315 1 1 0
I 15 3
315 2
1 0 1 /6 1 4 0
0 0 I/5 2 2
116 4 1 0 315 2 0 0
315 2 0 0 515 0 0 0
318
015
4 0
2 3 0 0 016 2 4 0 0
315 2 0 0 0 315 2 0 0 0 215 2
5/8
2 1 0 418 3 0 1 518 2 1 0 5/6 1 0 0
1
0 018
5 2 1
0 218 4 1 1 0
016
3 3 0 0
0
1 0 0
5/6 013 6/12 215 315 1 2 3 3 0 0 3 0 0 0 0 0 0 0 0 0 0 0 0 0 pre = before transplantation, aGVH,S - =during acute GVHD without corticosteroid treatment, aGVH,S + = during acute GVHD with corticosteroid treatment, cGVH = during chronic GVHD, noGVH =during quiescent period without GVHD. Biopsy histology indicates the histopathological findings in the epithelium graded as: 0 = no changes, I =vacuolar degeneration of basal cells, I1 =vacuolar changes with spongiosis and dyskeratosis, I11 = epidermolysis and bulla formation. The number of positive cells in immunohistochemical stainings is expressed as follows: - =not detectable or only occasional, + = slightly increased numbers, =moderate infiltrates, + = intense infiltrates, and + =very large infiltrates. CD57+ cells
++ +
+ ++ +++ ++++
++
++
1117
LESKINEN el al.
histology was normal in 6/23 sections. Other biopsies displayed grade 1-111 changes. Immunohistological stainings were done from 20 biopsies for CD2, CD4 and CD8 antibodies. In the biopsies taken in the absence of corticosteroid treatment, epidermal CD2 + , CD4 + and CD8 lymphocytes were seen in 10/ 12 biopsies. The number of CD 57-positive cells was increased in the epidermis in three biopsies. In the dermis, CD2+ and CD4+ infiltrates with intensity grades of 1 to + 4 were seen in all biopsies, and CD8 infiltrates in 10/ 12 biopsies. In all but two biopsies the number of all these types of lymphocytes was increased compared to biopsies taken before BMT. CD57positive cells were seen in the dermis in 6/12 biopsies. Corticosteroid treatment effectively influenced the number of both epidermal and dermal lymphocytes; most infiltrates were grade I + in severity. The results of detailed examination of the biopsies taken from cutaneous lesions during aGVHD, which displayed epidermal changes typical for aGVHD and where immunohistological stainings for most subclasses were available, are given in Table 2. The lymphoid infiltrates were more intense in the dermis than in the epidermis, but in most biopsies the same subclasses of mononuclear cells were seen in both
+ +
Day Post BMT +I2 +25 +49
+57 $34 +22 $77 +28 +28 +I9 +13 +10 +24 +53
Histology I I1 I I11 I 111
I1 I1 111
I1
I I1
I1 I11
structures. In the epidermis, FMC 33-positive cells were seen of grade 1 to 2+ infiltrates, but in the dermis they formed large infiltrates.
+
Biopsies taken during cG VHD ( n = 7) The histology was normal in all cases, except for parakeratosis noticed in one sample. In spite of this, CD4+ or CD8 T cells were found in all biopsies both in the epidermis and dermis. The epidermal and dermal clusters of cells consisted mostly of CD4+ lymphocytes, but the number of CD2+, CD8+ and CD57f lymphocytes was also increased in most biopsies when compared with pretransplant biopsies. The dermal infiltrates were graded + 1 to + 2 in intensity.
+
Biopsies taken during periods of no clinical GVHD (n=5) Three biopsies were histologically normal, one was classified as grade I GVHD (taken at autopsy) and one as grade I1 GVHD (the patient had suffered from a mild skin GVHD three weeks before). CD4+ lymphocytes ( + 1 to + 2 in intensity) were seen in the epidermis in 415 samples, CD8 + and CD2 + cells ( + 1 in intensity) in only two samples. In the dermis, CD2 , CD4+ and CD8+ cells were seen in slightly increased numbers only around the capillaries.
+
TABLE 2. Biopsies from skin lesions during aGVHD Epidermis Dermis CD2 CD4 CD8 CD57 FMC33 CD2 CD4 CD8
-
+ ++ ++ + + ++ ++ +++ ++- +
-
+ ++ ++ ++ ++
-
++ ++ ++ ++ + + ++ + + + + ++ + + +- ++++ ++ -
+
-
+ + ++ + + ++ + + ++ + -
-
+
-
-
+
+ ++++ +++ +++ + ++ +++ +- +
+
+++ ++ ++ ++ +++ +++ ++ ++ ++ ++ + +++ +++ + ++
-
+++ +++ +++ + +++ +++ ++ ++ ++ + + +++ +
CD57 FMC33 CS
-
+ ++ ++ ++ -
-
+-
++ +++ ++ ++ +++ ++
-
-
+ + +++ + +
Histology indicates epidermal histopathological changes: I: vacuolar degeneration of basal cells, 11: vacuolar changes together with spongiosis and dyskeratosis, 111: epidermolysis and bulla formation. The numbers of positive cells are expressed with symbols: - =not detectable or only occasional, + =slightly increased numbers, + + =moderate infiltrates, =large infiltrates, + + + + =very intense infiltrates. CS = corticosteroid treatment at the time of biopsy.
+++
1118
CUTANEOUS AND RECTAL IMMUNOHISTOLOGY IN GVHD
TABLE 3. Immunohistological findings in rectal biopsies of BMT patients Biopsy histology Patient groups Pre aGVH,S aGVH,S cGVH noGVH 0 61 10 315 419 1I2 214 1 4 2 3 0 2 11 0 0 0 1 0 111 0 0 1 0 0 IV 0 0 1 0 0 Epithelium CD2+ cells 318 1 I5 217 012 1 I4 5 2 4 1 3 0 2 1 1 0 0 0 0 0 0 CD4+ cells 619 115 217 012 214 2 4 4 2 2 1 0 0 0 0 0 0 0 1 0 CD8+ cells 1I 9 015 017 012 1/3 6 3 3 0 2 2 2 3 1 0 1 1 0 0 0 CD57+ cells 419 214 1 i6 oi2 0/4 5 2 4 2 4 0 0 1 0 0 0 0 0 0 0 Lamina propria CD2+ cells 018 014 018 012 014 6 2 2 0 3 5 0 1 2 2 0 0 I 1 0 0 0 0 1 0 CD4+ cells 019 015 018 012 014 5 I 1 0 2 6 0 2 3 4 1 0 0 1 0 0 0 0 2 0 CD8+ cells 119 015 012 013 018 7 2 3 0 1 4 0 2 1 2 0 1 1 2 0 0 0 0 0 0 CD57+ cells 719 013 215 012 1/ 4 2 3 3 2 3 0 0 0 0 0 0 0 0 0 0 ++++ 0 0 0 0 0 Biopsy histology indicates the histopathological findings in gut epithelium. 0 =no changes, I = focal dilatation and degeneration of mucosal glands, I1 = focal or diffuse loss of mucosal glands, 111= mucosal denudation. The patient groups and the coding of the intensity of cellular infiltrates are the same as in Table I .
+
+ ++ +++ + +++ ++
+ ++
+++
+ ++ +++
+ ++ +++ ++++
+ ++ +++ ++++
+ ++ +++ ++++
+ ++ +++
1119
LESKINEN
CD20+ B cells were not found in any of the skin biopsies. Rectal biopsies The results of rectal biopsies are summarized in Tables 3 and 4 Pretransplant biopsies ( n = 10) The histology was mostly normal, but slight dilatation of the crypts was seen in four biopsies. CD2 +, CD8 + and CD57 + lymphocytes were seen in the epithelium in most biopsies. No or only occasional CD4+ cells were seen in the epithelium in 6/9 biopsies. In the lamina propria CD2 +, CD4 + and CD8 + cells were detected in practically all biopsies stained with these antibodies. In one sample CD4+ cells were increased to grade 3 + and in three samples to 2 + infiltrates, while CD8+ cells of grade 2 + intensity were seen in one sample. These patients had been given cytostatic drugs only some days or weeks before the biopsy. Biopsies taken during aG VHD ( n = 14) Slight dilatation of crypts, coded as I, was seen in five biopsies. In one biopsy epithelial denudation, grade 111, and in one biopsy grade IV was seen. In spite of the slight epithelial changes, the total number of lymphocytes in the lamina propria was increased in 11/ 14 biopsies. In the glandular epithelium there was no clear difference in the numbers of CD2-, CD8- and CD57-positive lymphocytes as compared to the pretransplant biopsies; instead, CD4 + lymphocytes were seen in small infiltrates in 9/ 12 biopsies. In the lamina propria the intensity of CD4+ cells increased up to 2 + in 1 1 / 13 biopsies, and the differences in the intensities of CD2+ and CD8+ infiltrates were even more clear compared to pretransplant biopsies. The corticosteroid treatment did not influence the changes in rectal biopsies as markedly as in skin biopsies. Biopsies taken during chronic GVHD ( n = 2) In the epithelium the number of CD8+ cells was increased. In the lamina propria the number of CD4 + cells exceeded that of CD8 cells and the infiltrates were more intensive than during aGVHD, up to grades + 3 and +4. In these two biopsies the strong lymphocytic infiltrate also reached the muscularis mucosae layer.
+
1120
et a1
Biopsies taken during no clinical G VHD ( n = 4 ) These findings did not differ from the pretransplant biopsies. No CD20+ B cells were detected in the rectal biopsies. The detailed results from rectal biopsies taken during acute GVHD are summarized in Table 4.
DISCUSSION We studied skin and rectum biopsies taken from BMT patients before transplantation, during GVHD, and during quiescent periods. The cellular infiltrates both in cutaneous and intestinal structures were seen considerably more clearly in immunohistological than in the usual histological stainings. The immunohistopathological changes are often local and therefore we did not count the cell numbers per area, but estimated the cell numbers in the whole biopsy and graded the findings from - to 4 + . This study demonstrates that typical features of the maculopapular skin lesions of aGVHD are infiltrates of both CD4+ and CD8+ lymphocytes and FMC33+ monocytes both in the dermis and the epidermis. The changes in rectal biopsies are more difficult to interpret because also the pretransplant biopsies contained a considerable number of both CD4+ and CD8 + cells. Moreover, most of the patients were receiving corticosteroid treatment at the time of the rectum biopsy, which could have diminished the intensity of cellular infiltrates. During intestinal GVHD the increase in the CD2 + and CD8 + infiltrates in the lamina propria compared to the pretransplant biopsies was the most typical finding. No marked differences in the epithelial infiltrates of CD8 + or CD57 + cells were seen during GVHD compared to the pre-BMT biopsies; instead, the numbers of CD4 + cells were slightly increased. Our results resemble those of Elliot et al. who found an increase in CD2 + , CD4 + and CD8 + lymphocytes in skin biopsies exhibiting aGVHD changes (2), but are in disagreement with some earlier studies where the lymphoid cells in cutaneous GVHD are reported to be predominantly CD8+ (4, 5, 6 , 7, 10). In some previous studies the antibody used for CD4+ cells was
CUTANEOUS AND RECTAL IMMUNOHISTOLOGY IN GVHD
Day post BMT 69 56 89 50 20 28 40 53 27 28
Histology
+ + + + ++ + ++ +
111
0 0 0 I 0 I I 0 0 0 IV I
+118
+ 123 + 110
TABLE 4. Rectal biousies taken durina aG V H D Epithelium Lamina propria CD2 CD4 CD8 CD57 CD2 CD4 -
-
+ + ++ ++ ++ ++ + + * +
+
-
++ ++ + ++ -
+ * +
+
+ + +
++ ++
-
++ ++ + ++ ++ ++ ++ * +++
+
++ -
+ + +
*
+
+ ++ + ++ ++ ++ ++ + ++ + +++ ++ ++ + ++ + + ++ ++ ++ +++ +++ ++ ++
CD8
CD57
CS
++ ++ + ++ ++ +++ ++ + + +++ ++ +++
+ +
+ + + + + +
-
-
++ + + +
-
-
-n -
- if
Histology indicates epithelial histopathological changes coded as: 0 =no changes, I = focal dilatation and degeneration of mucosal glands, I1 =focal or diffuse loss of intestinal glands, 111= mucosal denudation, IV = diffuse mucosal denudation. The number of positive cells is graded as in Table 2 CS = corticosteroid treatment * =epithelial denudation these patients had skin and liver aGVHD, the others clinical skin and gut aGVHD. I'
OKT4 (Ortho) ( 5 , 6, 7), whereas we have used T4 (Coulter) because of the better antibody function in the immunoperoxidase staining. The staining methods in these earlier studies were also different, the immunofluorescence method usually being employed (2, 6, 7). The increase in the number of monocytes and NK cells in skin GVHD in our study is in agreement with Elliot et al. (2). There are only a few reports of the immunohistology of rectal biopsies during aGVHD (1, 11). Dilly & Sloane reported an increase in the number of CD8+ lymphocytes in the lamina propria with no change in CD4 + cells ( I ) . Weisdorf et al. found increased numbers of CD8 , as well as CD57+ lymphocytes, both in the epithelium and lamina propria during GVHD (11). We could not observe any increase in CD57 lymphoid cells during GVHD either in the gut epithelium or in the lamina propria. Most of the CD57+ cells seen in our biopsies in every patient group were obviously the neuroendocrine cells of the gut epithelium. Otherwise our results are concordant with the two earlier reports, and perhaps the different prophylaxis for GVHD in these studies could have influenced the results. The immunohistology of skin biopsies might be useful in the differential diagnosis of skin
+
+
problems after BMT because it is difficult to see the lymphoid cell infiltrates in normal histology stainings. On the other hand, only epidermal histological changes are frequently seen after BMT. Total body irradiation of 1000 rad causes very mild and transient epithelial atypia, which vanishes within a week. Cyclophosphamide and total body irradiation together can produce dyskeratosis with or without satellitosis, but these changes disappear during the three weeks after transplantation. The epithelial damage caused by chemotherapy and irradiation cannot be distinguished from the epithelial changes caused by aGVHD (9). Therefore the diagnosis of aGVHD should also include dermal lymphoid cell infiltration (3, 9). The cellular composition of the infiltrate in the dermis in aGVHD is typical and immunohistological stainings can help in the diagnosis of atypical skin lesions of aGVHD. The rectal biopsy histology in intestinal aGVHD is unspecific and the changes are often local. The pathogenesis of intestinal symptoms during aGVHD is also more complicated than in the skin. The initial intestinal lesion caused by effector T cells can facilitate the invasion of Gram-negative bacteria into the lamina propria, and endotoxin from these bacteria could enhance the immunological reaction by stimulat-
1121
LESKINEN er (11.
ing the release of interleukin-1, interferon and tumour necrosis factor. Thus the intensity and quality of the cellular infiltrates of the rectal mucosa may not have as clear a correlation with the clinical GVHD symptoms as in the skin. In immunohistological stainings the increase in CD8+ lymphocytes in the lamina propria during aGVHD is a constant though not always prominent finding. In patients whose intestinal symptoms are severe enough to need treatment and where the GVHD diagnosis is uncertain, rectal biopsy with immunohistological stainings may prove to be of some help, although information on the cellular infiltrates in infectious gut problems, which constitute the main differential diagnosis, is so far very limited.
REFERENCES Dilly, S. A . & Sloane, J . I?: Changes in rectal leucocytes after allogeneic bone marrow transplantation. Clin. Exp. Immunol. 67: 151-158, 1987. Elliot, C. J., Sloune, J. II, Pallett, G. D.& Sundemon, K. V: Cutaneous leucocyte composition after human allogeneic bone marrow transplantation: relationship to marrow purging, histology and clinical rash. Histopathology 12: 1-16, 1988. Elliot, C. J., Sloane, J. I?, Sanderson, K. V , Vincent, M . , Shepherd, V & Powles, R.: The histological diagnosis of cutaneous graft versus host disease: relationship of skin changes to marrow purging and other clinical variables. Histopathology 11: 145-155, 1987. Guyotat. D., Mauduit, G., Chouvet, B., Kanitakis,
I122
5.
6.
7.
8.
9.
10.
11.
J., Vu Van, H . , Fiere, D. & Thivolet, V: A sequential study of histological changes in the skin after allogenic bone marrow transplantation. Transplantation 41: 340-342, 1986. Janossy, G., Montano, L., Selby, W S., Duke, O., Panayi, G., Lampert, I., Thomas, J. A . . Granger, S., Bofll, M., Tidman, N . , Thomas, N. C. & Goldstein, G.: T cell abnormalities in tissue lesions developing during autoimmune disorders, viral infection and graft-versus-host disease. J. Clin. Immunol. (Suppl.); 2: 42-56, 1982. Kuye, V N., Neumann, I? M., Kersey, J., Goltz, R. W , Baldridge, B. D.. Michael, A . E & Platt. J. L.: Identity of immune cells in graft-versushost disease of the skin. Am. J. Pathol. 116: 436440, 1984. Lampert, I. A , , Janossy, G., Suitters, A . J., Janossy, G., Thomas, J. A., Palmer, S. & Gordon Smith, E.; Immunological analysis of the skin in graft versus host disease. Clin. Exp. Immunol. SO: 123-131, 1982. Lerner, K. G., Kao, G. E , Storb, R., Buckner, C. D., Clift, R. A . & Thomas, E. D.: Histopathology of graft-vs.-host reaction (GvHR) in human recipients of marrow from HL-A-matched sibling donors. Transplant. Proc. 6: 367-371, 1974. Sale, G. E., Lerner, K. G.. Barker, E. A , , Shulmun, H. M . & Thomas, E. D.: The skin biopsy in the diagnosis of acute graft-versus-host disease in man. Am. J. Pathol. 89: 621-634, 1977. Sloane, J. I?, Thomas, J. A., Imrie, S. F , Easton. D. E & Powless, R. L.: Morphological and immunohistological changes in the skin in allogeneic bone marrow recipients. J. Clin. Pathol. 37: 919-930, 1984. Weisclorf; S. A.. Roy, J., Snover, D.,Platt, J. L. & Weisdor- D.J.: Inflammatory cells in graftversus-host disease on the rectum: immunopathologic analysis. Bone Marrow Transplant. 7: 297-301, 1991.
