Diagnosis of cytomegalovirus (CMV) infection in pediatric transplant patients by the antigenemia, shell vial, and conventional culture assays performed on blood: correlation with CMV disease

Diagnostic ELSEVIER

Clinical and Diagnostic Virology 6 (1996) 51 61

Virology

Diagnosis of cytomegalovirus (CMV) infection in pediatric transplant patients by the antigenemia, shell vial, and conventional culture assays performed on blood: correlation with CMV disease Louise Pedneault a,u,c,*, Margaret Anglow ~,c, Carolina Alfieri a,bx, Earl Rubin l'b aDepartment of Microbiology and Immunology, Faculty of Medicine, Universit~ de Montrbal, Montreal, Quebec, Canada bDepartment of Microbiology and Infectious Diseases, H6pital Sainte-Justine, Montrkal, Qukbec, Canada CResearch Center, H6pital Sainte-Justine, Montrdal, Qubbec, Canada Received 11 August 1995; revised 5 January 1996; accepted 9 January 1996

Abstract Background: Human cytomegalovirus (CMV) is a significant cause of morbidity and mortality in transplant recipients. Isolation of CMV from blood leukocytes (CMV viremia) is considered predictive of CMV disease in transplant recipients. Therefore, investigation of methods for the rapid detection of CMV in the blood is important for diagnosis and management of these patients. Objective: To compare three techniques for the diagnosis and monitoring of CMV infection in a pediatric transplant population through the quantitative detection of CMV in peripheral blood leukocytes (PBL). Methods: Serial blood specimens were obtained for most patients. After separation of the PBL from each specimen, aliquots of the PBL were used for direct detection of CMV antigenemia by immunoperoxidase staining of acetone-fixed cells (CMV-vue kit, INCSTAR), and by immunofluorescence staining of formaldehyde-fixed cells (Complete 1C3 kit, Biosoft Argene). PBL were also inoculated into conventional cell culture tubes and shell vials. Patients' medical records were reviewed to ascertain the clinical significance of the results. Results: A total of 154 specimens obtained from 38 pediatric transplant recipients were evaluated. CMV was detected in 16 specimens obtained from eight patients: 11 specimens were found positive with the CMV-vue kit, 10 with the Complete 1C3 kit, four by conventional culture, and one by the shell vial assay. Seven of the eight patients with CMV-positive PBL had clinical signs and other laboratory evidence of active CMV infection. In general, a high-level antigenemia was demonstrated in the presence of clinical disease, but there were exceptions. Conclusions: The two antigenemia kits were more sensitive than conventional culture and the shell vial assay for the detection of CMV in the blood of pediatric transplant patients. Our results suggest that CMV antigenemia is a sensitive and specific rapid method for the diagnosis and monitoring of CMV infection in our patient population.

Keywords: Cytomegalovirus; Pediatric transplant recipients; Antigenemia; Shell vial; Culture Abbreviations: CMV, cytomegalovirus; PBL, peripheral blood leukocytes; D, donor; R, recipient; lg, immunoglobulin; +, positive; - , negative; ind, indeterminate; PBS, phosphate-buffered saline; CPE, cytopathic effect; CF, complement fixation; PCR, polymerase

chain reaction. * Corresponding author. Bristol-Myers Squibb Pharmaceutical Research Institute, 5 Research Parkway, Dept. 102, Wallingford, CT 06492, USA. Fax: + 1 203 2846852; Tel.: + 1 203 2847757. t Present address: Department of Microbiology and Infectious Diseases, Montreal Children's Hospital, Montrral, Qurbec, Canada.

0928-0197/96/$15.00 © 1996 Elsevier Science B.V. All rights reserved PII S0928-0197(96)00205-X

52

L. Pedneault et al. / ~Tinica/ and Diagnostic Virology 6 (1996) 51 61

1. Introduction

Human cytomegalovirus (CMV) remains a significant cause of morbidity and mortality in transplant recipients. It is often very difficult to differentiate the clinical manifestations of active CMV infection from allograft rejection or from infection with other opportunistic microorganisms. Appropriate matching of blood and graft donors with the recipient regarding CMV serostatus should prevent most primary CMV infections. This approach would be especially beneficial to pediatric transplant recipients, since most of them are still CMV-seronegative at the time of transplantation. Unfortunately, the shortage of CMV-seronegative organ donors renders the application of such measures rather impractical. Prophylactic regimens, including an antiviral drug and/or CMV hyperimmune globulins, have been shown to either prevent or reduce the severity of primary CMV infections. Furthermore, antiviral drugs, with or without the adjunct of CMV hyperimmune globulins, are also currently used as therapy when needed. Consequently, it is extremely important to correctly detect CMV for appropriate diagnosis and monitoring of infection, but also for monitoring of antiviral activity. Isolation of CMV from blood leukocytes (CMV viremia) is considered predictive of CMV disease in transplant patients. It appears to correlate better with significant disease than CMV isolation from urine or throat specimens (Meyers et al., 1990). Conventional culture methods are of limited usefulness because of the slow growth of the virus and because of frequent toxicity problems observed when blood specimens are inoculated into cell culture. Shell vial cultures were shown by several studies either to be less sensitive than conventional cultures for detection of CMV viremia or to be negative in a fair percentage of CMV positive blood cultures (Paya et al., 1987; Wunderli et al., 1989; Erice et al., 1992; Mazzulli et al., 1993). More recently, a new method (CMV antigenemia assay) has been developed to detect the lower matrix structural protein of CMV, namely pp65, directly in peripheral blood leukocytes (PBL) based on immunological staining using monoclonal antibodies. Several groups have

already shown that CMV antigenemia is a very rapid and sensitive technique (van der Bij et al., 1988; van den Berg et al., 1989; Gerna et al., 1990; Gerna et al., 1991; van den Berg et al., 1991; Boland et al., 1992; Erice et al., 1992; Bitsch et al., 1993; Gerna et al., 1993; Koskinen et al., 1993; Landry and Ferguson, 1993; Mazzulli et al., 1993; Perez et al., 1994; Storch et al., 1994; Weber et al., 1994; Erice et al., 1995; Meyer-Konig et al., 1995). Two kits are now commercially available for the detection of CMV antigenemia: one relies on immunoperoxidase staining of acetone-fixed cells (CMV-vue kit; INCSTAR Corp., Stillwater, MN), and the other on immunofluorescence staining of formaldehyde-fixed cells (Complete 1C3 kit; Biosoft Argene, Varilhes, France). In the present study, we used the CMV-vue and Complete 1C3 kits for the quantitative detection of CMV antigenemia and we compared the results with those obtained using the shell vial assay and conventional cell culture for detecting CMV viremia in sequential blood specimens from pediatric transplant patients. Results were correlated with clinical symptoms and antiviral therapy.

2. Materials and methods

2.1. Patient population Between March 1993 and March 1995, 154 blood specimens were collected from 38 pediatric transplant recipients seen at Sainte-Justine Hospital. A minimum of three blood samples were serially obtained from 26 transplanted children (22 liver, two kidney, one liver and small bowel, and one bone marrow). This was part of routine surveillance studies during the first 3 months postengraftment or, when clinical symptoms compatible with a CMV infection were noted, this was also used for diagnosis and follow-up of such infection, and for monitoring of antiviral therapy. Four more transplanted children (two liver, one liver and bowel, and one heart) were tested twice, and eight others (four kidney, two heart, one liver, and one bowel) were evaluated once. The CMV serostatus of donors (D) and recipients (R) was determined by the detection of CMV

L. Pedneault et al. / Clinical and Diagnostic Virology 6 (1996) 51-61

immunoglobulin G (IgG), as described in Section 2.5. Such CMV carrier status was defined as positive ( + ) , negative ( - ) or, in cases where the information was not available, as indeterminate (ind). Prophylactic regimens against CMV infection varied according to the type of organ being transplanted and to the CMV serostatus of the donor and recipient. Earlier on in our study, D + / R liver transplant patients received CMV hyperimmune globulin with or without ganciclovir, as part of another research protocol. When this other research project was over, this same subgroup of patients was given only CMV hyperimmune globulin as prophylaxis. One D + / R + liver transplant recipient (patient 8) received intravenous gammaglobulins for prophylaxis. CMV hyperimmune globulin prophylaxis was also routinely administered to D + / R renal transplant recipients. Lastly, the only bone marrow recipient (D + / R + ) received acyclovir for 30 days as prophylaxis. Rejection episodes in liver transplant patients were documented by percutaneous liver biopsy and were treated with high doses of steroids (intravenous methylprednisolone) or, in cases of steroid-resistant rejection, with a course of OKT3 monoclonal antibody therapy.

2.2. Collection and separation of leukocytes All blood specimens were processed within 2 h of collection. PBLs were separated from 3-5 ml of heparinized blood by gradient centrifugation using the Polymorphprep solution (Nycomed Pharma AS, Oslo, Norway) and washed once with phosphate-buffered saline (PBS). They were then resuspended in PBS and counted. Aliquots of PBLs from each specimen were washed once in minimum essential medium (Gibco, Grand Island, NY), and resuspended in the same medium. Their concentration was adjusted to 2 x 106/ml, and they were set aside for inoculation of conventional cell culture tubes and shell vials. The remaining PBLs from each specimen were washed again in PBS and submitted to a 5-min incubation in a 0.8% NH4CI solution (reagent included in the Complete 1C3 kit) at room temperature, for lysis

53

of contaminating erythrocytes. After another wash, PBLs were resuspended in PBS, and their concentration was adjusted to 2 x 106/ml for testing by the two CMV antigenemia kits.

2.3. C M V antigenemia assays The CMV-vue and Complete 1C3 CMV antigenemia kits were used following the manufacturer's instructions, except for a single modification in the CMV-vue assay and two changes in the Complete 1C3 kit procedure. For the CMV-vue assay, 5 x 104 cells were applied to each of four wells of a microscopic glass slide. Coverslips were placed over the cell suspensions and incubated at room temperature for 15 min to allow the cells to sediment in an even distribution. Coverslips were removed, and slides were then fixed in acetone, soaked in a solution that inhibits endogenous peroxidase, incubated for 45 min at 37°C with a cocktail of two monoclonal antibodies (clones C10 and C11) directed at the pp65 of CMV, and stained for 45 min at 37°C using a horseradish peroxidase labeled anti-mouse IgG antibody. Slides were counterstained with hematoxylin and observed under a light microscope at 400 × magnification. The manufacturer recommended reading two wells seeded as described above. However, in this study, two (1 x 105 cells) and, whenever possible, four (2 × 105 cells) wells were read and the results of the two readings were compared. Red-brown nuclear or perinuclear staining of the PBLs was considered positive. The results obtained by reading two wells (according to the supplier's recommendations) were used in Table 1 for comparison with the Complete 1C3 kit. The slides provided by the manufacturer contained one positive and one negative control well each, consisting of CMV-infected and uninfected fibroblasts, respectively. For the Complete 1C3 assay, 2 x 105 cells were cytocentrifuged onto three glass slides, fixed with a formaldehyde solution, and permeabilized with the Nonidet P-40 solution. We then added an incubation step in goat serum (Gibco) diluted 1:5 in PBS for 20 min at 37°C, to eliminate nonspecific background fluorescence. According to the manufacturer's instructions, one slide was then

L. Pedneault et al./ Clinical and Diagm)stic Virology 6 (1996) 51 61

54

Table l C o m p a r i s o n o f the C M V - v u e a n d C o m p l e t e 1C3 a n t i g e n e m i a a s s a y s with the shell vial t e c h n i q u e a n d c o n v e n t i o n a l cell c u l t u r e f o r d e t e c t i o n o f C M V in 154 p e r i p h e r a l b l o o d l e u k o c y t e s p e c i m e n s C M V - v u e (1 x cells")

10 s

C o m p l e t e 1C3 (2 × l0 5 cells ~')

Shell vial (7 × 11)5 cells h)

Cell c u l t u r e (7 × l0 s cellsbl

Number of specimens

+

+

+ + -

+ +

Not done + Not done -

+ + ~ + -

1 1 I 1 5 5 2 121 17

k -

Not done

" N u m b e r o f p e r i p h e r a l b l o o d l e u k o c y t e s used f o r e a c h a n t i g e n e m i a assay, a c c o r d i n g to the m a n u f a c t u r e r ' s i n s t r u c t i o n s . bTotal n u m b e r o f p e r i p h e r a l b l o o d l e u k o c y t e s (2 x 105 cells + 5 × II) 5 cells, see text) used for the shell vial a s s a y a n d c o n v e n t i o n a l culture.

incubated with a monoclonal antibody directed against the pp65 of CMV (clone 1C3) for 30 min at 37°C. As control for specificity, we chose to incubate the second slide with mouse serum (Organon Teknica, Scarborough, Ontario, Canada) diluted 1:250 in PBS, and the third slide with a mixture of monoclonal antibodies directed against an immediate early antigen (72 kDa) of CMV (kind gift of Dr B. Brodeur, Unit6 de Recherche en Vaccinologie, Croix-Rouge/ Recherche et D6veloppement, Ste-Foy, Qu6bec, Canada (Brodeur et al., 1992)) for 30 min at 37°C. All slides were then stained with an antimouse fluorescein isothiocyanate conjugate for 30 min at 37°C. Slides were examined under oil immersion, at 1000 × magnification, using a Zeiss microscope equipped with epifluorescence. A specimen was considered positive for CMV if one or more typically fluorescent polylobate nuclei were observed. 2.4, C M V viremia

For detection of infectious CMV in blood, both conventional cell culture and the shell vial assay (Gieaves et al., 1985) were performed. Amounts of 2 x 105 and 5 x 105 PBL were inoculated into each of two cell culture tubes and two shell vials in which human foreskin fibroblasts had been grown for 2 - 4 days. Once inoculated, each

vial was spun at 700 x g for 1 h at room temperature. Excess inoculum was removed, and the shell vials were refed with culture medium (50% minimum essential medium (Gibco), 50% 199 medium (Gibco), 2% fetal calf serum, and antibiotics). After an incubation period of 24 h for the vial inoculated with 5 x 105 cells and of 48 h for the vial containing 2 x 105 cells, coverslips in each vial were fixed in cold acetone, stained by indirect immunofluorescence using the mixture of monoclonal antibodies directed against an immediate early antigen (72 kDa) of CMV (kind gift of Dr B. Brodeur (Brodeur et al., 1992)), and examined for nuclear fluorescence typical of CMV-infected fibroblasts. The cell culture tubes were maintained in culture medium for 6 weeks, whenever possible, and regularly monitored for the appearance of a CMV cytopathic effect (CPE). Suspicion of a CMV-specific CPE was confirmed by indirect immunofluorescence performed on the trypsinized cells from the presumably infected tube, using the mixture of monoclonal antibodies directed against the CMV 72-kDa protein. 2.5. Other virologic studies

Other specimens collected from our transplant patients for viral culture were processed according to standard procedures (Schmidt, 1989). CMV

L. Pedneault et al. / Clinical and Diagnostic Virology 6 (1996) 51-61

serology was performed by an enzyme immunoassay following the supplier's recommendations (Enzygnost, Behringwerke, Marburg, Germany), for the detection of specific IgG and IgM antibodies, and by complement fixation (CF) to evaluate paired sera for a fourfold antibody titer increase (Schmidt and Emmons, 1989). 2.6. Definitions

Diagnosis of active CMV infection was documented by positive cultures of blood, urine, throat or organ biopsy and/or by a notable serological event. Primary infection was defined as the first appearance of a CMV positive culture or of CMV IgM antibodies in a previously CMV-seronegative transplant recipient. Secondary infection (reactivation or reinfection) was diagnosed by a fourfold increase in CMV antibody titers, detection of CMV IgM and/or virus isolation in a CMVseropositive patient. Symptomatic CMV infection was termed severe when active CMV infection was accompanied by two or more of the following CMV-related manifestations without evidence of other causes: unexplained fever for 3 or more days, thrombocytopenia, liver enzyme elevations ('transaminitis'), and clinical and/or culture evidence of organ involvement (CMV hepatitis, and gastrointestinal ulceration such as duodenitis or esophagitis). Symptomatic CMV infection that did not fulfill these criteria was termed mild.

3. Results

A total of 154 PBL specimens obtained from 38 pediatric transplant recipients were evaluated. Table 1 summarizes the results obtained following comparison of the two antigenemia kits with the shell vial assay and conventional culture for the detection of CMV. The virus was identified by one or more methods in 16/154 (10.4%) specimens obtained from eight patients. The CMV-vue and the Complete IC3 CMV antigenemia kits allowed detection of the virus in 11 and 10 of the 16 positive specimens, respectively. At least one of the two antigenemia assays was positive in 15 of

55

the 16 CMV-positive specimens. All positive and negative controls included in the performance of the two antigenemia assays gave the expected results. Conventional culture allowed identification of four CMV isolates, including one strain missed by the other techniques. CMV was identified in only one specimen by the shell vial assay (also culture-positive), and quantitation revealed only two positive cells out of a total of 7 x 105 cells inoculated into two shell vials. However, the shell vial assay was not performed on two occasions when conventional culture was found positive. Eleven cultures were interrupted due to toxicity (n = 9) or bacterial contamination (n = 2), and one shell vial was uninterpretable because of toxicity. These 12 tests were entered as negative in Table 1. Seven of the eight patients with CMV-positive PBL by antigenemia had clinical signs and other laboratory evidence of active CMV infection (Table 2). Patient 1 received transfusions prior to her graft and, therefore, her CMV status was indeterminate. She developed a severe primary CMV infection, manifested by hepatitis. Of interest, when using 2 x l05 cells to perform the CMVvue assay, the PBL specimen collected on day 26, when the patient only had mild fever, was already positive for the presence of CMV. Patient 2 was a CMV-seronegative 2-year-old boy who received a CMV-seropositive liver graft. He was febrile for 2 days before transplantation, for which a complete bacterial and viral work-up was performed and no underlying cause could be identified. The antigenemia assay was positive 5 days post-transplant, and severe rejection was diagnosed histopathologically on the liver biopsy. The patient was treated only for rejection and no specific antiviral therapy was given. Later on, the primary CMV infection was also documented by the appearance of CMV IgM and CMV-positive blood, urine and throat cultures. Primary CMV infection was diagnosed 1 month post-liver transplant in patient 3. During the 3 years which followed his transplantation, urine and throat cultures were continuously positive for the presence of CMV, and CMV IgM antibodies

L. Pedneault et al. / Clin&al and Diagnostic Virology 6 (1996) 51 61

56 Table 2

C o r r e l a t i o n between clinical a n d virological findings in eight t r a n s p l a n t recipients with p e r i p h e r a l b l o o d l e u k o c y t e specimens positive for C M V Patient/age CMV immune Time after Antigenemia at T x status Tx (years)/Tx (donor/recipient) Complete 1C3 CMV-vue type (2x105 cellsb) ( l x l 0 scells b) I / 16/liver

+/indeterminatec

4

0d

0

0

0 0

0 26

3 84

Not done 0

43

50

0

34

2

+

48

72

30

96

0

Cont

56

8

3

14

0

-

Not done 0

Not done 0

Not done -

3/I/liver

+/-

+/-

Not done 10

Not done 0

0

75

0

0

(I

0

-

5

0

4

14

0

-

11

39

4

18

0

-

18 122

0 0

0 0

0 0

0 Not done

+

2,42 ~

0

0

0

0

3,4 d

0

4

8

0

3,6"

2

0

0

0

4/7/bone marrow

+ /+

3,1if' 33 weeks

0 0

0 2

0 Not done

0 0

5/2/liver

?/+

7,19 ~'

0

3

12

0

7,21 ''

0

0

0

0

6/10/kidney +:

55

5

(1

7/12/liver

76 39

+/-

0 Not done

0 Not done

Comments

CMV-vue (2x105cells b)

26 34

59 62

2/2/liver

Shell vial Culture (7 × 105 cells b) (7 x 105 cells b)

-

Not done

0

+

Not done Not done

0 Not done

Not done

Mild fever Mild fever, mild 'transaminitis' Moderate rejection, hepatitis, high dose steroids, 2 C M V + urine cultures Hepatitis, C M V + liver BX and urine cultures Moderate fever, mild hepatitis GCV therapy started Mild fever, mild hepatitis End of GCV treatment Severe rejection, moderate fever, high dose steroids Moderate rejection and fever C M V + urine and throat cultures, CMV IgM + Mild rejection, high dose steroids, CMV + urine culture Thrombocytopenia, petechial rash, mild 'transaminitis', CMV lgM + Mild 'transaminitis'. mild liver fibrosis Duodenitis. esophagitis, ascitis, hepatomegaly, moderate fever, C M V + duodenum and urine cultures GI tract hemorrhage, aorto-duodenal fistula Mild 'transaminitis,' CMV IgM equivocal Moderate fever, dry cough CMV IgM + Severe duodenitis, moderate fever, C M V + duodenum culture

L. Pedneault et al. / Clinical and Diagnostic Virology 6 (1996) 51-61

57

Table 2 (contd.) Patient/age CMV immune Time after Antigenemia Shell vial Culture Comments (7 x l0 s cellsb) (7 x 105 cellsb) status Tx at Tx Complete 1C3 C M V - v u e CMV-vue (years)/Tx (donor/recipient) (2 x 105 censb) (1 × 10s cellsb) (2 × 10s cellsb) type

8/12/liver

+/+

45

134

26

Not done

Not done

+

~1

0

0

Not done

0

Toxic

35 42

150 206

40 85

Not done Not done

0 0

-

47

0

0

Not done

0

-

Moderate fever, dry cough, CMV IgM+, GCV therapy started Moderate fever, URTI, influenza A and parainfluenza 3 + nasopharyngeal culture, CMV+ urine culture, IVIG GCV therapy started Severe esophagitis, moderate fever High fever,

Aspergillus fumigatus 66

Not done

Not done

Not done

Not done

Not done

perisplenic abscess CMV IgM+, significant increase in CMV CF antibodies

T~, transplantation; CMV Ig, CMV hyperimmune globulins; cont, contaminated; Bx, biopsy; GCV, ganciclovir; GI, gastrointestinal; PCR, polymerase chain reaction; IVIG, intravenous gammaglobulins; URTI, upper respiratory tract infection; CF, complement fixation. Walues in years, weeks. bNumber of peripheral blood leukocytes used for each assay. cPatient had received blood products unscreened for CMV 7 days before Tx. dNumber of CMV-positive cells.

were frequently found, suggesting that the patient may have been undergoing chronic CMV reactivation. More recently, he received high doses of steroids for mild rejection, at a time when his urine culture grew CMV. Fourteen weeks later, the patient developed thrombocytopenia and a petechial rash, accompanied by a mild 'transaminitis'. He did not receive antiviral treatment and his evolution was satisfactory. In the months to follow, the patient suffered from a series of severe bacterial infections, and all blood cultures and antigenemia testings for CMV remained negative. Primary CMV infection occurred 26 days prior to bone marrow transplantation in patient 4. Thirty-three weeks after transplantation, he developed severe secondary CMV infection for which he received therapy with ganciclovir and CMV hyperimmune globulin. Seven years post-liver transplantation, patient 5 had a gastrointestinal haemorrhage, as a result of duodenal ulceration. The CMV antigenemia was positive, however the conventional and shell vial

cultures of the duodenum biopsy were negative. Massive transfusions were administered with blood products not screened for CMV. Shortly after, the patient started mounting a CMV IgM response, while a mild and persistent 'transaminitis' was observed. The polymerase chain reaction (PCR) for the detection of hepatitis C genome was performed on this patient's blood and yielded a positive result (data not shown). Patient 6 developed a mild primary CMV infection post-kidney transplantation. Post-liver transplant primary CMV infection also occurred in patient 7, and presented as a severe duodenitis accompanied by moderate fever. Therapy with ganciclovir was administered. The patient later developed a clinical syndrome compatible with an influenza infection, the blood viral cultures and the antigenemia assays remained negative. Following treatment of several episodes of rejection with high dose steroids and OKT3, patient 8, a liver transplant recipient, developed severe sec-

58

L. Pedneault et al. / Clinical and Diagnostic Virology 6 (1996) 51 61

ondary CMV infection, manifested by severe esophagitis accompanied by fever. Finally, CMV antigenemia assays and blood viral cultures were consistently negative in five transplant patients (three kidney, one liver and one heart) who developed primary CMV infection (documented by a positive CMV IgM serology and/or a CMV-positive urine culture) around the time at which the blood assays were performed. Four of these patients were completely asymptomatic and one had only fever (data not shown).

4. Discussion

In the present study, we compared the CMV-vue and Complete 1C3 CMV antigenemia kits with the shell vial assay and conventional culture for the diagnosis and monitoring of active CMV infection in a population composed strictly of pediatric transplant recipients. In accordance with others who focused on populations predominantly composed of adult individuals receiving an organ transplant or infected with HIV, we have shown the CMV antigenemia assay to be an accurate and early marker of active CMV infection in pediatric transplant recipients as well. As previously reported, CMV antigenemia was generally found to be more sensitive than the shell vial assay and/or conventional culture (Gerna et al., 1990; van den Berg et al., 1991; Boeckh et al., 1992; Erice et al., 1992; The et al., 1992; Landry and Ferguson, 1993; Mazzulli et al., 1993; Erice et al., 1995), to yield a positive result at an earlier stage of active CMV infection compared to CMV viremia detection methods (van den Berg et al., 1991; Boeckh et al., 1992), to precede the appearance of serological markers of active CMV infection (van der Bij et al., 1988; van den Berg et al., 1989; van den Berg et al., 1991; The et al., 1992; Bitsch et al., 1993), and to remain positive while blood viral cultures were already negative during the first days of ganciclovir therapy (Gerna et al., 1991; Boeckh et al., 1992; Erice et al., 1992; Landry and Ferguson, 1993). The period when the CMV antigenemia detection was positive corresponded with the symptomatic period of active CMV infection. Storch et al. (1994) found the shell vial and

antigenemia assays to be of equivalent sensitivity for the detection of CMV. They suggested that the differences they observed in the relative sensitivities of the antigenemia and shell vial assays, in comparison to previous reports, were the result of their use of a quantitative modification of the shell vial assay. Since we also used the shell vial assay in a quantitative fashion and since, in our hands, these assays were still less sensitive than both antigenemia assays, this modification cannot be the only factor responsible for the discrepancy observed between the antigenemia and viremia methods. Furthermore, they used the CMV-vue assay for the detection of CMV antigenemia, and found the high degree of technical skill required for accurate reading of the slides to be a major disadvantage, which could perhaps partly explain their results. Lastly, because of procedural differences, it is difficult to compare our study with the study of Storch et al. (1994). These technical differences include the number of PBLs inoculated into the shell vials, the types of cell cultures and the monoclonal antibodies used to perform the shell vial procedure. When comparing the two technologies for the detection of CMV antigenemia, we found that the Complete 1C3 kit resulted in clearly superior cell morphology, brighter nuclear staining, and less artifacts compared to the CMV-vue kit. The appearance of a CMV-positive cell with the Complete 1C3 kit was very characteristic, rendering the reading of the slides much easier. The main parameters of immunostaining techniques were evaluated by several investigators. They all found better readability when using formaldehyde instead of acetone or methanol-acetone fixation, and immunofluorescence rather than immunoperoxidase staining (Gerna et al., 1992; Landry and Ferguson, 1993; Boeckh et al., 1994; Perez et al., 1995). These results are in agreement with our findings, since formaldehyde fixation and immunofluorescence are used in the Complete 1C3 kit, and since the CMV-vue kit is based on the use of acetone fixation followed by immunoperoxidase staining. In our hands, the total number of positive samples was equivalent with both assays. The discordant results between the two assays were observed mostly when the samples examined contained a small number of positive cells. Some of

L. Pedneault et al. / Clinical and Diagnostic Virology 6 (1996) 51-61

these differences can probably be attributed to sampling variability. The fact that a single monoclonal antibody was used in the Complete 1C3 kit, versus a pool of two monoclonal antibodies in the CMV-vue kit may also contribute to the discrepancy. However, when the number of positive cells in a specimen was elevated, the Complete 1C3 assay would allow a higher number of positive cells to be detected. The use of a different number of cells in each assay and the easier readability of formaldehyde fixation and immunofluorescence staining can probably account for at least some of these latter discrepancies. At least 32 (33 if patient 1, D + / R ind. is included) of the 38 patients were CMV-seronegative at the time of transplantation, as would be expected in a pediatric transplant population. As previously reported, appropriate matching of blood and graft donors with the recipient regarding CMV serostatus did prevent primary CMV infection in at least 13 CMV-seronegative children who received a transplant from a CMV-seronegative donor. Of the eight patients who developed CMV antigenemia, four had primary CMV infection (patients 1, 2, 6 and 7) and three others had secondary CMV infection (patients 3, 4 and 8). Among the patients who developed primary CMV infection, two liver transplant recipients had a severe CMV clinical syndrome (patients 1 and 7), and one patient each with a liver (patient 2) and a kidney (patient 6) transplant suffered only from mild symptoms (fever). The three patients who were afflicted with secondary CMV infection all had severe symptoms, and two o f them developed their symptomatic episode following treatment of rejection with high dose steroids with or without OKT3 (patients 3 and 8). Patient 4 developed primary CMV infection immediately prior to transplant. Although we considered this to be secondary CMV infection, it is likely that the severe nature of the disease is due to the proximity of CMV infection just prior to intense immunosuppression. In patient 5, the aorto-duodenal fistula and secondary gastrointestinal haemorrhage most likely resulted from a complicated duodenal ulceration. Active CMV infection could not be

59

demonstrated at the time of the acute episode. However, it is possible that reactivation of CMV infection in this immunosuppressed patient may have been the cause of the initial duodenal ulceration. CMV IgM did become detectable shortly after this acute episode. This IgM response may be due to reactivated CMV disease. However, the blood products received were not CMV screened. Therefore, an IgM response due to CMV acquisition from blood products cannot be excluded. As previously reported (van der Bij et al., 1988; van den Berg et al., 1989; Gerna et al., 1990; Boeckh et al., 1992; The et al., 1992; Landry and Ferguson, 1993; Mazzulli et al., 1993), we also have found that, in general, high levels of antigenemia are associated with severe clinical symptoms of CMV infection; however, discrepancies were observed. In our experience and that of others (van den Berg et al., 1989; Boeckh et al., 1992; The et al., 1992; Koskinen et al., 1993; Landry and Ferguson, 1993; MazzuUi et al., 1993), some patients had a high number of antigen-positive cells associated with minor symptoms, whereas others suffered from a severe CMV syndrome with only a few antigenpositive cells suggesting that host factors probably also play a role in the expression of disease symptoms. The choice of a cut-off level in the CMV antigenemia assay to differentiate between clinically unimportant and significant CMV infections has been highly variable, depending on the definitions of CMV clinical syndromes, the transplant population studied, the number of cells used per slide and the immunostaining technique selected: > 10 positive cells/0.75 x 105 (van der Bij et al., 1988) or /1.5 × 105 cells (van den Berg et al., 1989) in renal transplants, > 50 (Gerna et al., 1991) or > 80 positive cells/2 x 105 cells (Gerna et al., 1990) and > 100 positive cells/0.5 × 105 cells (Koskinen et al., 1993) in heart transplants, and > 1 positive cell/0.5 x 105 cells in bone marrow transplants (Boeckh et al., 1992). In our study, using the number of cells per slide recommended by the manufacturer for each CMV antigenemia assay, < 5 positive cells may or may not have been associated with clini-

60

L. Pedneault et al. / Clinical and Diagnostic Virology 6 (1996) 51 61

cal symptoms but, by immunofluorescence, > 50 positive cells always correlated with severe clinical symptoms. In one bone marrow recipient (patient 4 in Table 2), we also found that the detection of very few positive cells (2/1 × 105 cells) was associated with severe CMV disease. In conclusion, the two antigenemia kits were more sensitive than conventional culture and the shell vial assay for the detection of CMV in the blood of pediatric transplant recipients. Both antigenemia assays are labor intensive, but the reading of the slides was found easier when using the Complete 1C3 kit. Our results suggest that CMV antigenemia is a sensitive and specific rapid method for the diagnosis and monitoring of active CMV infection in our patient population. In general, a high-level antigenemia was demonstrated in the presence of clinical disease. However, there were exceptions, and the quantitative results obtained did not always correlate well with the severity of CMV infection. At last, due to the inherent difficulties in obtaining large amounts of blood from a venipuncture in pediatric patients, CMV antigenemia assays might have a definite advantage in this patient population over the methods involving a culture step.

Acknowledgements We are indebted to the gastroenterology, nephrology, and cardiac transplant teams for their collaboration in this study. Appreciation is extended to Argene S.A. and INCSTAR Corp. for supplying the CMV antigenemia kits, to the staff of the Hrpital Sainte-Justine's Diagnostic Virology Laboratory for their excellent technical assistance, and to Mary Cassidy for expert secretarial assistance. L.P. was a clinical research scholar supported in part by the Fonds de la Recherche en Sant6 du Qu6bec. M.A. was a fellow of the Programme Canadien de Bourses de la Francophonie. C.A. is a research scholar of the J.A. De Srve Foundation. The authors have no financial interest in any of the diagnostic companies mentioned in this report.

References Bitsch, A., Kirchner, H., Dennin, R., Hoyer, J., Fricke, L., Steinhoff, J., Sack, K. and Bein, G. (1993) The long persistence of CMV DNA in the blood of renal transplant patients after recovery from CMV infection. Transplantation, 56, 108-113. Boeckh, M., Bowden, R.A., Goodrich, J.M., Pettinger, M. and Meyers, J.D. (1992) Cytomegalovirus antigen detection in peripheral blood leukocytes after allogenic marrow transplantation. Blood, 80, 1358-1364. Boeckh, M., Woogerd, P.M., Stevens-Ayers, T., Ray, C.G. and Bowden R.A. (1994) Factors influencing detection of quantitative cytomegalovirus antigenemia. J. Clin. Microbiol., 32, 832 834. Boland, G.J., de Weger, R.A., Tilanus, M.G.J., Ververs, C., Bosboom-Kalsbeek, K. and de Gast, G.C. (1992) Detection of cytomegalovirus (CMV) in granulocytes by polymerase chain reaction compared with the CMV antigen test. J. Clin. Microbiol., 30, 1763-1767. Brodeur, B.R., Lussier, M., Larose, Y., Rossier, E., Miller, H., Nakagawa, C. and Evelegh, M. (1992) New monoclonal antibodies for the detection of immediate early antigens of cytomegalovirus. Viral lmmunol., 5, 61-69. Erice, A., Holm, M.A., Gill, P.C., Henry, S., Dirksen, C.L., Dunn, D.L., Hillam, R.P. and Balfour, H.H. Jr. (1992) Cytomegalovirus (CMV) antigenemia assay is more sensitive than shell vial cultures for rapid detection of CMV in polymorphonuclear blood leukocytes. J. Clin. Microbiol., 30, 2822 2825. Erice, A., Holm, M.A., Sanjuan, M.V., Dunn, D.L., Gill, P.C. and Balfour, H.H. Jr. (1995) Evaluation of CMV-vue antigenemia assay for rapid detection of cytomegalovirus in mixed-leukocyte blood fractions. J. Clin. Microbiol., 33, 1014 1015. Gerna, G., Revello, M.G., Percivalle, E., Zavattoni, M., Parea, M. and Battaglia, M. (1990) Quantification of human cytomegalovirus viremia by using monoclonal antibodies to different viral proteins. J. Clin. Microbiol., 28, 2681 2688. Gerna, G., Zipeto, D., Parea, M., Revello, M.G., Silini, E., Percivalle, E., Zavattoni, M., Grossi, P. and Milanesi, G. (1991) Monitoring of human cytomegalovirus infections and ganciclovir treatment in heart transplant recipients by determination of viremia, antigenemia, and DNAemia. J. Infect. Dis., 164, 488-498. Gerna, G., Revello, M.G., Percivalle, E. and Morini, F. (1992) Comparison of different immunostaining techniques and monoclonal antibodies to the lower matrix phosphoprotein (pp65) for optimal quantitation of human cytomegalovirus antigenemia. J. Clin. Microbiol., 30, 1232-1237. Gerna, G., Percivalle, E., Revello, M.G. and Morini, F. (1993) Correlation of quantitative human cytomegalovirus pp65-, p72- and pl50-antigenemia, viremia and circulating endothelial giant cells with clinical symptoms and antiviral treatment in immunocompromised patients. Clin. Diagn. Virol., 1, 47 59. Gleaves, C.A., Smith, T.F., Shuster, E.A. and Pearson, G.R. (1985) Comparison of standard tube and shell vial cell

L. Pedneault et al. / Clinical and Diagnostic Virology 6 (1996) 51-61

culture techniques for the detection of cytomegalovirus in clinical specimens. J. Clin. Microbiol., 21, 217-221. Koskinen, P.K., Nieminen, M.S., Mattila, S.P., Hayry, P.J. and Lautenschlager, I.T. (1993) The correlation between symptomatic CMV infection and CMV antigenemia in heart allograft recipients. Transplantation, 55, 547-551. Landry, M.L. and Ferguson, D. (1993) Comparison of quantitative cytomegalovirus antigenemia assay with culture methods and correlation with clinical disease. J. Clin. Microbiol., 31, 2851-2856. Mazzulli, T., Rubin, R.H., Ferraro, M.J., D'Aquila, R.T., Doveikis, S.A., Smith, B.R., The, T.H. and Hirsch, M.S. (1993) Cytomegalovirus antigenemia: clinical correlations in transplant recipients and in persons with AIDS. J. Clin. Microbiol., 31, 2824-2827. Meyer-Konig, U., Serr, A., von Laer, D., Kirste, G., Wolff, C., Haller, O., Neumann-Haefelin, D. and Hufert, F.T. (1995) Human cytomegalovirus immediate early and late transcripts in peripheral blood leukocytes: diagnostic value in renal transplant recipients. J. Infect. Dis., 171,705-709. Meyers, J.D., Ljungman, P. and Fisher, L.D. (1990) Cytomegalovirus excretion as a predictor of cytomegalovirus disease after marrow transplantation: importance of cytomegalovirus viremia. J. Infect. Dis., 162, 373-380. Paya, C.V., Wold, A.D. and Smith, T.F. (1987) Detection of cytomegalovirus infections in specimens other than urine by the shell vial assay and conventional tube cell cultures. J. Clin. Microbiol., 25, 755-757. Perez, J.L., Mariscal, D., Tubau, F., Niubo, J. and Martin, R. (1994) Evaluation of the CMV-vue antigenemia assay for rapid detection of cytomegalovirus in peripheral blood leukocytes. Diagn. Microbiol. Infect. Dis., 19, 15-18. Perez, J.L., De Ona, M., Niubo, J., Villar, H., Melon, S., Garcia, A. and Martin, R. (1995) Comparison of several fixation methods for cytomegalovirus antigenemia assay. J. Clin. Microbiol., 33, 1646-1649. Schmidt, N.J. (1989) Cell culture procedures for diagnostic virology. In: N.J. Schmidt and R.W. Emmons (Eds), Diagnostic Procedures for Viral, Rickettsial and Chlamydial Infections, American Public Health Association, Washington, DC, pp. 51-100. Schmidt, N.J. and Emmons, R.W. (1989) General principles of laboratory diagnostic methods for viral, rickettsial and

61

chlamydial infections. In: N.J. Schmidt and R.W. Emmons (Eds), Diagnostic Procedures for Viral, Rickettsial and Chlamydial Infections, American Public Health Association, Washington, DC, pp. 21-28. Storch, G.A., Buller, R.S., Bailey, T.C., Ettinger, N.A., Langlois, T., Gaudreault-Keener, M. and Welby, P.L. (1994) Comparison of PCR and pp65 antigenemia assay with quantitative shell vial culture for detection of cytomegalovirus in blood leukocytes from solid-organ transplant recipients. J. Clin. Microbiol., 32, 997-1003. The, T.H., van der Ploeg, M., van den Berg, A.P., Vlieger, A.M., van der Giessen, M. and van Son, W.J. (1992) Direct detection of cytomegalovirus in peripheral blood leukocytes - - a review of the antigenemia assay and polymerase chain reaction. Transplantation, 54, 193-198. van den Berg, A.P., van der Bij, W., van Son, W.J., Anema, J., van der Giessen, M., Schirm, J., Tegzess, A.M. and The, T.H. (1989) Cytomegalovirus antigenemia as a useful marker of symptomatic cytomegalovirus infection after renal transplantation - - a report of 130 consecutive patients. Transplantation, 48, 991-995. van den Berg, A.P., Klompmaker, I.J., Haagsma, E.B., Scholten-Sampson, A., Bijleveld, C.M.A., Schirm, J., van der Giessen, M., Slooff, M.J.H. and The, T.H. (1991) Antigenemia in the diagnosis and monitoring of active cytomegalovirus infection after liver transplantation. J. Infect. Dis., 164, 265-270. van der Bij, W., Torensma, R., Van Son, W.J., Anema, J., Schirm, J., Tegzess, A.M. and The, T.H. (1988) Rapid immunodiagnosis of active cytomegalovirus infection by monoclonal antibody staining of blood leucocytes. J. Med. Virol., 25, 179 188. Weber, B., Nestler, U., Ernst, W., Rabenau, H., Braner, J., Birkenbach, A., Scheuermann, E.-H., Schoeppe, W. and Doerr, H.W. (1994) Low correlation of human cytomegalovirus DNA amplification by polymerase chain reaction with cytomegalovirus disease in organ transplant recipients. J. Med. Virol., 43, 187-193. Wunderli, W., Kagi, M.K., Gruter, E. and Auracher, J.D. (1989) Detection of cytomegalovirus in peripheral leukocytes by different methods. J. Clin. Microbiol., 27, 19161917.