Differential Diagnosis of Acute Allograft Rejection and CMV-Infection in Renal Transplantation by Urinary Cytology

Path. Res. Pract. 180, 161-168 (1985)

Differential Diagnosis of Acute Allograft Rejection and CMVInfection in Renal Transplantation by Urinary Cytology Martin Winkelmann, Bernd Grabensee and Peter Pfitzer Dept. of Internal Medicine A (Chairman: Prof. Dr. W. Schneider) and Division of Cytopathology (Director: Prof. Dr. P. Pfitzer) of the Institute of Pathology (Chairman: Prof. Dr. W. Hort) University of Dusseldorf

SUMMARY Acute allograft rejection and CMV-infection are the most common complications after renal transplantation. Quick differential diagnosis between these two complications is still difficult but necessary, since both complications demand a different therapy. More than 2500 urinary samples from 33 transplanted patients were prospectively examined and part of them evaluated quantitatively. Urinary samples of patients with acute renal failure, long-term haemodialysis or immunosuppressive therapy served as controls. The following cytomorphological criteria proved to be useful: Tubular epithelial cells, casts, oxalate crystals (sand-glass shaped), dirty background, increasing erythrocyturia, mixed cell clusters, lymphocytes and mitoses. Rejection is going on when the number of renal tubular cells is increased and two or more further criteria are positive. 25 acute allograft rejections without acute renal failure were diagnosed clinically. All 25 rejections were also diagnosed by urinary cytology. Nevertheless, it is not possible to differentiate between acute allograft rejection and acute renal failure of other origin. CMV-infection was serologically detected in 7 patients. In 6 of them viral infected cells were found in the urine shortly after the onset of unspecific clinical symptoms. Besides the typical "owl-eye" cells milkglass nuclei, sometimes with eosinophilic condensation, were seen while criteria for transplant rejection were never observed at the same time. Cytologic examination of voided urine is a simple diagnostic help for the differentiation between allograft rejection and CMV-infection after renal transplantation.

Introduction Renal transplantation has become a routine therapy for endstage renal failure? The most common complications are acute allograft rejection and CMV-infection. Both require quick diagnosis because therapy is fundamentally different. Diagnosis of acute allograft rejection is mainly based on the clinical criteria: unexplained fever, reduction in urine flow with malaise, increase in blood pressure, tenderness and enlargement of the graft completed by laboratory data on increasing creatinine and BUN. However, only some patients show the complete picture with all symptoms and moreover these symptoms can also be © 1985 by Gustav Fischer Verlag, Stuttgart

observed in connection with other complications after renal transplantation. For instance CMV-infection is occurring in more than 50% of allograft recipients?' 18. CMV has also primarily to be diagnosed clinically since serologic or cultural identification cannot be expected before 10 to 14 days after onset. As a consequence of the overlapping symptoms differentiation between acute allograft rejection and CMVinfection is not always possible. The histologic diagnosis takes at least two days and does not guarantee clear results in any case, but risks occasional bleeding complications. Investigators having tried to detect cytologic criteria for acute allograft rejection l , 3, 6,11,15, 1',21, 2!', put quite 0344-0338/8510180-0161$3.50/0

162 . M. Winkelmann, B. Grabensee and P. Pfitzer

different emphasis on the morphologic findings. It was the aim of our investigation to find in a prospective study cytomorphological criteria which allow to diagnose acute renal allograft rejection and to differentiate between rejection and CMV-infection.

more than 2500 urinary samples were investigated. As control served: 6 patients with end stage renal failure treated with haemodialysis and diuresis of at least 500 ml per day,S patients with acute renal failure after shock, 3 patients with skin diseases and normal renal function under immunosuppressive therapy (150 mg azathioprine and 10-30 mg a-methyl-prednisolone). For preparation freshly voided urine is put 20 minutes into the refrigerator (4°) for spontaneous sedimentation. Then 8 ml of the sediment are centrifuged for 5 minutes at 1500 rpm. 3 to 5 gt of the sediment are used for the cytocentrifuge (Cytospin/SHANDON) and 3 slides per sample are prepared. Two of them are stained according to Papanicolaou and one to May-GriinwaldGiemsa. One of the slides is evaluated quantitatively according to the following principles: General impression: cellularity, contaminations, state of preservation, dirty background (necrosis), amount of tubular epithelial cells, casts, mixed cell clusters, oxalate crystals, mitoses.

Materials and Methods Over a period of 20 months (1980-1982) 33 renal transplantations were carried out on 30 patients (16 males, 14 females) with an age range between 21 and 53 years. All recipients had been treated with chronic haemodialysis or continuous ambulatory peritoneal dialysis (CAPD) for more than 2 years. Patients with residual diuresis had at least one urine specimen examined before transplantation. During hospitalisation specimen were examined daily and after being discharged at every follow-up. Altogether,

, a

c

d

Fig. 1. Urinary cytology in case of allograft rejection a-b) casts, dirty background, erythrocytes and degenerated tubular cells Papanicolaou 250 x. c) casts and oxalate crystals in sand-glass form Papanicolaou 850 x. d) degenerated tubular cells with oxalate crystals i Papanicolaou 450 x.

Urinary Transplant Cytology . 163 Total amount in 5 fields (16 x): Erythrocytes, granulocytes, lymphocytes, macrophages, urothelial cells, not identified cells. Furthermore, the following clinical facts are taken into account: serum creatinine value, diuresis, clinical signs of acute renal failure, clinically diagnosed and treated acute allograft rejections, urinary tract infection, clinically suspected CMV-infection, intravenous application of contrast medium.

numerous (Fig. 1 a, c, 2 e-f). If there are no complications the cytological findings return to normal. Only some vital renal tubular cells are sometimes found even months after transplantation (Fig. 2 a-b).

Diagnosis of acute rejection crisis

Cytological picture after renal transplantation without complication.

In 33 transplantations 34 acute rejection crises were clinically diagnosed of which 10 were controlled and confirmed histologically. 25 of the acute rejection crises occurred when no concurrent acute renal failure was apparent. In these cases the following 8 cytological phenomena can be observed (Fig. 3):

During the first post-operative hours a macrohaematuria can be observed which is reduced to a microhaematuria during the following days and ceases approximately after 10 days. Besides detritus some casts (Fig. 1 a-c), sandglass shaped oxalate crystals (Fig. 1 c-d) and a slightly increased number of granulocytes are found during the first three days. During this period macrophages are more frequent than urothelial cells and lymphocytes while degenerative renal tubular cells are quite

Renal tubular cells (Fig. 2): In all 25 rejection crises an increased number of renal tubular cells, single or clusters, is observed. At least 3 clusters per slide or 7-10 single tubular cells in five fields are classified as increased. Two types of cells can be differentiated: Normal tubular cells of round or oval shape with eccentric nuclei (Fig. 2 a-c) and tubular cells with nuclear alterations such as prominent nucleoli, altered nuclear-cytoplasmic ratio and especially

Results

'"

)

c

Fig. 2. Tubular cells a-b) clusters of tubular cells, c-d) mitoses in tubular cells, e-f) beginning and advanced degeneration of tubular cells, a-b), d-f) Papanicolaou, c) MGG 450 x.

164 . M. Winkelmann, B. Grabensee and P. Pfitzer

1p

1r;

~lt_u_b_u_la_r_ep

__l_th_e_l_ia_l_c_e_lIs__________

ca __s_t_s____________________

I

degenerative alterations including pycnosis (Fig. 1 a, c, 2 e-f). The percentage of altered tubular cells increases when rejection progresses.

•

~)1000ro

-J)

~I

n

2,5

96

Casts (Fig. 1 a-c): In 24 cases (96%) a significant cylindruria with at least 10 casts per slide can be observed. Protein casts are the most frequent ones, sometimes granular and blood casts are found as well.

0 10

dirty background

Dirty background (Fig. 1 a-bY: In 16 cases (64%) cellular detritus, damaged nuclei and casts make a so-called dirty background which can easily be differentiated from inflammatory alterations or bacterial contamination.

(Increasing) erythrocyturia mixed cell clusters oxalate crystals C8andgl.shaped>

Erythrocytes: In 16 cases (64%) an increase in red blood cells is observed and in three of them even macrohaematuria. In some patients this is the first sign of acute rejection crisis when other causes for haematuria are excluded.

48 010

48~

Mixed cell clusters (Fig. 4a-b): They are found in 12 cases (48 %) and consist of detritus, pycnotic renal tubular cells and sometimes lymphocytes or granulocytes. Their number increases with acute rejection crisis.

I

lymph> 240f0

cyturla

I

mitos) 200f0

Oxalate crystals (Fig. 1 b-c): In 12 cases (48%) numerous sand-glass shaped oxalate crystals are found. They are

.-....

Fig. 3. Frequency of criteria during acute allograft rejection.

.,.' .." .'.. 1,'..• . ', ..,.•.:. .,

•

, •• # ' .

\ • ~.,t..'.'

.e ,

,

,

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

. d

'. •

a Fig. 4. a-b) Mixed cell clusters Papanicolaou 450 x.

b

.., .

• • •• • •

•

.,

" I , . : .' •••

.....

-.• ."• .

'

\

Urinary Transplant Cytology· 165

n

6 4

2 2

a

D 3

4

5

D 6

D

7

Criteria of 8

Rejection

Fig. 5. a) Number of criteria found during transplant rejection, b) Time of cytological diagnosis for allograft rejection.

Diagnosis of acute allograft rejection in patients with concurrent renal failure

-4

b

-3 -2

-1

,

0

In 14 of the 33 renal transplantations there was a postoperative acute renal failure and in 9 events acute allograft rejections were diagnosed clinically. Cytological examination was complicated in all these cases by severe leukocyturia and erythrocyturia. Nevertheless, 6 acute allograft rejections could be diagnosed cytologically. However, 7 cases without clinical symptoms had a cytological picture that fulfilled the criteria for acute rejection crisis. As a result there were 6 correct positive, 3 false negative and 7 false positive cytological diagnoses III patients with acute renal failure after transplantation.

1

2

-3--'--

Days

Clinical Diagnosis

very rare in the urine of patients with normal serum creatinine levels « 1.4 mg/%) and rare in patients without complications.

Lymphocytes: Contrary to earlier findings lymphocyturia is seen in our series in only 6 cases of acute rejection crisis (24%). Lymphoblasts are not observed. In one case, however, lymphocyturia started 10 days before the onset of the clinical symptoms of acute rejection crisis. Mitoses (Fig. 2 c-d): In 5 cases mitoses (20% are found in renal tubular cells. The quantitative analysis of the other urinary elements such as granulocytes, macrophages and urothelial cells show no correlation to acute rejection crisis. According to our evaluation acute rejection crisis is going on when there is an increase in renal tubular cells and when two or more of the other criteria are positive (Fig. 5 a-b). In 3 cases acute rejection was diagnosed cytologically without corresponding clinical symptoms. Only in one case the application of contrast medium could explain these cytologic changess. In the other two cases no clinical event could be found to explain the cytological diagnosis.

Controls Cytology in patients with acute renal failure after shock: In 5 controls with acute renal failure the cytological picture is the same as in acute allograft rejection. The changes are observed at the commencement of diuresis and at the end of the polyuric phase. All criteria (Fig. 3) are found including lymphocytes in one case. Cytology in patients with endstage renal failure for years: Usually a few granulocytes and red blood cells are found while renal tubular cells are present only sporadically. Casts can be seen but neither mitoses nor oxalate crystals. Cytology in patients under immunosuppressive therapy without impaired renal function: There are no specific findings and none of the rejection criteria.

CMV-infection Two types of viral infected cells are observed in Papanicolaou stained preparations. One type are renal tubular or squammous cells with so-called milkglass nuclei (Fig. 6 a-b) or some eosinophilic nuclear condensations. The second type are renal tubular cells with dense nuclear inclusions surrounded by a halo and a thickened nuclear membrane - the so-called owl-eye cells (Fig. 6 a-f). Multinuclear, probably herpes infected cells can sometimes be found. In six out of seven patients with a more than fourfold increase of CMV-titre after two weeks such viral infected

166 . M. Winkelmann, B. Grabensee and P. Pfitzer

a

b

I'

'~e

d

c cells had been seen two to six days after the onset of unspecific clinial symptoms. In four only milkglass nuclei and in the remaining two both alterations were present. Since the described cytological criteria of acute allograft rejection were lacking, no CMV-infection was misinterpreted cytologically as allograft rejection. Clinically, however, two CMV-infections were thought to be probably acute allograft rejections as raised serum creatinine levels were present. Under the application of an anti-rejection therapy the clinical symptoms persisted and renal function became worse (Fig. 7). Discussion Renal transplantation has become an established routine therapy for endstage renal failure. As the technical problems are almost solved, the function of the allograft depends mainly on immunological factors 14. By HLAserotyping the survival of the allograft has been prolonged but nevertheless losses are still due to acute or chronic rejection crises in 90% of the cases7.

f

... e

•

Fig. 6. CMV-manifestation after renal transplant. a-b) milk-glass nuclei Papanicolaou a) 450 x, b) 600 x, c-f) owl-eye nuclei Papanicolaou 600 x.

In the sixties first cytologic findings were published trying to diagnose acute rejection crisis by an increased number of lymphocytes in the urine 8, 11, 16,21. Later, other morphological alterations were found with the main emphasis on renal tubular cells3, 4, 6, 12, 15, 17, 19, 20, 22. Our series confirm these findings as we observed no rejection crisis without increase in renal tubular cells. Contrary to the earliest publications 8, 11, 16, 21 lymphocyturia was observed only in every forth case of acute allograft rejection. Furthermore, we saw lymphocyturia in one case without any clinical symptoms and in another the patient was suffering from acute renal failure lO • Identification of lymphocytes with monoclonal antibodies in fine needle aspiration cytology of the allograft could become a prom. ising diagnostic help24. Through renal tubular cells are the most important cell type we agree with other authors 3, 15, 17 that further criteria are necessary to be sure that an acute rejection crisis is going on. Casts are naturally not very specific but acute allograft rejection must be doubted when casts are missing because in 96% of all crises they are found. As far as red blood cells are concerned only a distinct increase is

Urinary Transplant Cytology· 167 of importance. In progressive acute allograft rejection the amount of red blood cells decreases again, presumably due to impaired local blood circulation. Mitoses in acute allograft rejection are up to now rarely described22 • We saw them in 20% of the acute rejection crises, in two cases of acute renal failure and in one case of pre-endstage renal failure. According to these observations mitoses can be observed when there is severely impaired renal function. Rather typical for quickly impairing renal function are sandglass shaped oxalate crystals25 , perhaps due to renal acidosis. It is, however, not clear why only sandglass shaped forms are seen instead of the usually more frequent square ones. All the described eight criteria are a sign for tubular lesions and according to our results they can be seen in acute renal failure as well. There is no cytologic symptom which is specific exclusively for acute allograft rejection. None of the described combinations of criteria appeared in the cases of CMV-infection. However, specific cytologic changes can be found occasionally as first reliable hint of a viral infection2, 9, 13, 15,20, 23. They can be seen in Papanicolaou-stained but never in Pappenheim-stained slides. Since milk-glass nuclei are observed more frequently than the owl-eye cells cytological differentiation between different viruses is not always possible. But according to general clinical experience only CMV-infection is in consideration. Generalized herpes infection is rare and infection with papovaviruses is neither of practical importance. Clinically, CMV-infection has obviously been underestimated l 8• Similar to the case mentioned, CMV-infection could be misinterpreted as acute allograft rejection. Therefore, cytologic examination of the urine is recommended when there is a clinical suspicion of CMV-

infection because high corticoid doses can be desastrous in acute CMV-infection. That is why we consider the differentiation between acute allograft rejection and CMVinfection an important cytologic task.

References I Atay Z, Zobl H, Georgii A (1973) Validitat der zytologischen Befunde im Urin nierentransplantierter Patienten. Mitt Ges Bek Krebskrankheit 6: 293-296 2 Bossen EH, Johnston J, Amatulli J, Rowlands DT (1969) Exfoliative zytopathologic studies in organ transplantation. I. The cytologic diagnosis of cytomegalic inclusion disease in the urine of renal allograft recipients. Am J. Clin Pathol52: 340-344 3 Bossen EH, Johnston WW, Amatulli J, Rowlands DT (1970) Exfoliative cytopathologic studies in organ transplantation. Cytologic profile of urine during acute renal allograft rejection. Acta Cytol14: 176-181 4 Bossen EH, Johnston WW (1977) Exfoliative cytopathologic studies in organ transplantation. V. The diagnosis of rejection in the immediate postoperative period. Acta Cytol21: 502-507 5 Cronin RE (1981) Acute renal failure with radiographic contrast media. Seminars in Nephrology 1: 51-55 6 Fidler JP, Dajani F, First MR, Munda R, Alexander JW (1978) Value of urine cytology in renal transplantation. Transplantation 26: 133-135 7 Grabensee B, Altrogge G ( 1980) Nierentransplantation. Intern Welt 5: 161-169 8 Hrushesky W, Sampson D, Murphy GP (1972) Lymphocyturia in human renal allograft rejection. Arch Surg 105: 424-426 9 Johnston WW, Bossen EH, Amatulli J, Rowlands DT (1969) Exfoliative cytopathologic studies in organ transplantation. II. Factors in the diagnosis of cytomegalic inclusion disease in urine of renal allograft recipients. Acta Cytol 13: 605-610

GPT64

oc

mgOfo

j

GPT50

j

GPT23

I

virocytes

a-methyl- 100

. . . ." " "________•

~n~~~

100

azathioprlne·:':"::l~~~~~~~L_ _ _ _ _~

Fig. 7. Clinical development of a case of CMV-infection after renal transplantation.

168 . M. Winkelmann, B. Grabensee and P. Pfitzer 10 Kaden J, Groth J (1980) Lymphozyturie nach Nierentransplantation. J Urol Nephr 73: 827-833 11 Kaufmann HM, Clark RF, Magee JH et al (1964) Lymphocytes in urine as an aid in the early detection of renal homograft rejection. Surg Gynecol Obstet 119: 25-36 12 Kline TS, Craighead JE (1967) Renal homotransplantation, the cytology of the urine sediment. Am J Clin Pathol47: 802-806 13 Koss LG, Coleman DV (1981) Advances in Clinical Cytology. Butterworth, London-Boston-Sydney-Durban-Wellington-Toronto 14 Lenhard V, Schafer A, Dreikorn K (1982) Bedeutung immunologischer Faktoren fur die klinische Nierentransplantation. Nieren- und Hochdruckkrankheiten 11: 25-35 15 O'Morchoe PJ, Erozan YS, Cooke CR et al (1976) Exfoliative cytology in the diagnosis of immunologic rejection in the transplanted kidney. Acta Cytol 20: 454-461 16 Papdimitriou M, Chrisholm GD, Kulatilake AE, Shackman R (1970) Clinical evaluation of the urinary sediment after renal allotransplantation. J Clin Pathol23: 99-103 17 Pellet H, Minaire E, Thonnerieux M (1981) Cytologie urinaire apres transplantation renale de diagnostic precoce de rejet. Arch Anat Cytol Pathol 29: 39-43

18 Richardson WP, Colvin RB, Cheeseman SH et al (1981) Glomerulopathy associated with cytomegalovirus viremia in renal allograft. New Engl J Med 305: 57-63 19 Schumann GB, Burleson RL, Henry JB, Jones DB (1977) Urinary cytodiagnosis of acute allograft rejection using the cytocentrifuge. Am J Clin Pathol 67: 134-140 20 Schumann GB, Burleson RL (1977) Importance of urine cytology in renal transplantation. Transplantation 23: 186-188 21 Spencer ES, Petersen VP (1967) The urinary sediment after renal transplantation. Acta Med Scand 182: 73-82 22 Taft PD, Flax MH (1966) Urinary cytology in renal transplantation: association of renal tubular cells and graft rejection. Transplantation 4: 194-204 23 Traystman MD, Gupta PK, Shah KV et al (1979) Identification of viruses in the urine of renal transplant recipients by cytomorphology. Acta Cytol 24: 501-510 24 Wood RFM, Bolton EM, Thompson JF, Morris PJ (1983) Monoclonal antibodies and fine needle aspiration cytology in detecting renal allograft rejection. Lancet July: 278 25 Zollinger HU, Mihatsch MJ (1978) Renal Pathology in Biopsy. Springer-Verlag, Berlin-Heidelberg-New York

Key words: Urinary cytology - Transplantation - Rejection - CMV-Infection Prof. Dr. Peter Pfitzer, Cytopathology, Moorenstr. 5, D-4000 Dusseldorf