The Cyfomegalovirus-Antigenemia Assay in the Diagnosis of ... - JASN

0 downloads 0 Views 1MB Size Report
In peripheral ..... culture results, positive histology. or detection of anti-CMV. 1gM. The ... defined as laboratory evidence of CMV infection together with fever of.
The Cyfomegalovirus-Antigenemia of Posttransplant Cytomegalovirus Brian

M. Murray,2

Kay

Jan

Schewegler,

BrentJens,3

Jocelyn

Daniel

P. Singh,

Amsterdam,

and

Rocco

J. Brentjens,

(J. Am.

Soc.

ABSTRACT Cytomegalovirus major

cause

1994;

(CMV)

and

continues

mortality

recipients, yet prompt A new assay has been

diagnosis developed

antigens

blood

to be a

in transplant

remains a problem. that detects CMV

leukocytes

(CMV-AG).

A retrospective analysis of the experience with this assay was performed, and its usefulness In the diagnosis

of CMV

infection

in renal

with unexplained fever conventional modalities tion

of circulating

results rapid

suggest and

sensitive

transplant

recipients

was compared with (buffy coat culture,

anti-CMV

that

immunoglobulin

the

CMV-AG

test

than

that of detecM). The

assay

existing

is a more

modalities

in

the early diagnosis of CMV infection. When expressed quantitatively, it can discriminate between CMV infection and CMV disease, and It Is useful in monitoring

the

course

of infection

and

the

response

to therapy. Key Words: nemia,

Cytomegaiovirus,

viral culture,

I

Recelv#{149}dAugust

2

Correspondence

Center, 3

renal

transplantation,

18. 1993.

Acc.pted

November

continues complication

Deceased.

of the American

Society

to be of

5. 1993.

to Dr. B.M. Murray, SUNY at Buffalo, 462 Grld#{149}r Street, Buffalo, NY 14215.

1046-6673/040816f5$03.00/0 Journal of the American SOCIetY of Nphrology CopyrIght C 1994 by the American Society

Journal

antige-

viral serology

ytomegalovirus (CMV) infection the most common infectious

C

Vanessa

Gray,

Irene

Pawlowski,

1615

Infection

In peripheral

Myers,

of Nephrology

4:1615-1622)

of morbidity

Jean

of Nephrology

J. Myers, Department of Microbiology, of New York at Buffalo, Buffalo, NY

Nephrol.

I

solid organ transplantation and results in considerabbe morbidity and occasional mortality (1). Early diagnosis of CMV infection remains an important goal, particularly with the development of new, promising antiviral therapy. Conventional diagnostic techniques involve the detection of serologic responses to CMV or the isolation of the virus from clinical specimens, especially buffy coat from peripheral blood. Virus is detected by its cytopathic effect on cultured cells or the detection of early CMV antigen expression in cultured cells by the shell vial technique with spin amplification (2). Even with the development of the shell vial technique. these tests arc time consuming, do not provide a sufficiently rapid diagnosis, and do not discriminate between patients with asymptomatic infection or invasive disease. Van der Bij et at. (3) first reported an immunocytobogic assay for the detection of CMV antigens (CMV-AG) in circulating peripheral blood leukocytes (PBL). They suggested that this might be a valuable tool for the early diagnosis and monitoring of active CMV disease in immunocompromised patients. By the use of a mixture of monocbonal antibodies, CMV bower matrix protein PP65 was detected in cytocentrifuged peripheral blood beukocytes (PBL) within 3 to 5 h after sampling. Results of this assay can be expressed quantitatively by reporting the number of CMV-AG-positive cells per 50,000 PBL, and the authors claimed that the CMV-AG assay was as sensitive as an anti-CMV ELISA (4). Indeed, CMV-AG could be detected earlier in the course of the illness, even before the onset of clinical symptoms (4). Similar results have been reported by some groups (5-7) but not by others (8,9). In all of these studies, patients were monitored prospectively with monthly or biweekly CMV diagnostic tests, which is a relatively costly approach to clinical management. The CMV-AG assay became available to us in late 1 990, and we elected to use it in a different manner. Patients were not prospectiveby monitored, but instead, CMV-AG assays were done in those patients in whom CMV infection was suspected. This study represents a retrospective analysis of a 20-month experience (October 1 990 to July 1992) with the CMV-AG assay in the diagnosis and monitoring of CMV infection in recipients of renal albografts. The utility of the antigen test was compared with that of more conventional diagnostic modalities. Our results support the concept that the CMV-AG assay will be an important addition to the diagnostic armamentarium used to detect CMV infection.

I. Pawlowski, K. Schewegler, Department State University of New York at Buffalo,

of Pathology, Buffalo, NY

in the Diagnosis

C. Venuto

B.M. Murray, V. Gray, J.P. Singh, P.C. Venuto, Departments of Medicine, Microbiology, and Pathology, State University of New York at Buffalo, and the Departments of Medicine and Virology, Erie County Medical Center, Buffalo, NY D. Amsterdam, State University

Assay

Erie County

Medical

CMV-Antigenemia

Assay

METHODS Initial

Study

Patients were renal transplant recipients who were transplanted between October 1 990 and July 1992 at one of two sites: Buffalo, New York-Erie County Medical Center (Site 1) or Buffalo General Hospital (Site 2). At Site 1 immunosuppression consisted of sequential quadruple therapy beginning with Mmnesota antilymphocyte globulin, methylprednisobone, and azathioprine. Cycbosporine was begun when the serum creatinine fell below 265 tmol/L (3 mg/dL), and the antilymphocyte globulin was discontinued once therapeutic cyclosporine bevels were achieved. The patients were then maintained on triplc immunosuppression for at least 1 yr. Site 2 used a similar triple-drug maintenance immunosuppression program but differed in that OKT3 and not Minnesota antilymphocyte globulin was used in the immediate posttransplant period (induction phase) and only in patients who experienced initial nonfunction. Patients who were seronegative for CMV before transplantation and who were given an organ from a seropositive donor received prophylaxis with immune globulin. Charts of all patients on whom CMV-AG assays were performed were reviewed. In general. CMV-AG assays were performed when dinical circumstances suggested the possibility of CMV infection, mostly in patients with unexplained fevers, and samples from both sites were processed in the same reference laboratory. Serial tests were occasionally obtained in CMV-AG-positive patients to monitor the course of infection. Charts were also reviewed to see if other diagnostic tests for CMV (blood and urine cultures, serology) were performed. ,

Culture

Techniques

PBL were isolated with Polymorph Prep (NycomedPharma, Oslo, Norway). Urine and isolated PBL were inoculated into MRC-5 cells (Biowhittaker, Walkersville, MD), and the presence of CMV was detected either by cytopathic effect in conventional cell monolayers or by detection of CMV early antigen in shell vials by the spin amplification technique. The CMV-AG assay was identical to that described by Van der BIj et at. (3) PBL were isolated from EDTAtreated venous blood by sedimentation with 5% Dcxtran Ficoll. The isolated PBL were washed three times in phosphate-buffered saline, resuspended, and cytocentrifuged onto glass slides. The slides were incubated with a mixture of two monocbonal antibodies to CMV bower matrix protein PP65, washed, and subsequently incubated with an appropriate dilution of peroxidase-labeled rabbit anti-mouse immunogbobubin. The enzyme reaction was carried out with 3amino-8-cthylcarbazolc substrate, followed by mild counter staining with hematoxylin and mounting in

1616

glycerol gelatin. The slides were examined with a light microscope at x250 magnification, and the number of CMV-AG-positive cells per 50,000 PBL was reported. A series of tests on blood from 10 normal healthy individuals and 6 transplant recipients drawn at the time of transplant showed no evidence of detectable CMV antigens by this techniquc.

Serology CMV immunoglobulin by the use of an assay (CMV CAP-M;

(1gM) levels enzyme-linked Biowhittaker).

were determined antibody capture

Definitions A laboratory diagnosis of CMV infection was made on the basis of positive buffy coat culture results, positive histology. or detection of anti-CMV 1gM. The presence of CMV disease was defined as laboratory evidence of CMV infection together with fever of 38.3#{176}Cor higher for 2 or more days within a 7-day period plus one or more of the following features: leukopenia (10

test in the diagnosis

OF

infection

and

disease

Positive

Specificity

in renal

Negative

(%)

(%)

Predictive Value (%)

Positive

100

90

95

100

Positive

100 100

64 93

74 93

100 100

100

93

93

100

PeakCMV-AG>40

COMPARISON

of CMV

ANTIGENEMIA

AND

BUFFY

COAT

Serial cuv-1’c

CULTURE

counts

in four

Predictive Value (%)

patients

with

CMV disease

350

300 a,

F

1

____

Buffy coat 14 positive

Buffy coat 13 negative

Buffy coat 23 negative

CMV

CMV

CMV

Ag

count

56±20

Figure 2. Comparison with simultaneous

Ag

count

12±10

buffy

blood

antigen

Day8

levels

Doy3O

in four

Day 60

patients

CMV

with

test

cultures.

the illness. Figure 3 shows serial CMV-AG counts in four patients with CMV disease who had at least weekly determinations of CMV-AG after the initial test was positive. The CMV-AG level rose to a peak approximately 1 wk after the initial diagnosis and then fell steadily over the next 60 days. Persistent low bevels of antigenemia were often noted up to 2 months after the initial presentation. In general. the peak of the CMV-AG level correlated with the intensity of disease, as shown in Figure 4 for a patient with predominantly hepatitis. Figure 5 shows the case of a 37-year-old diabetic, seronegative for CMV pretransplant. who received a cadaver kidney from a seropositive donor and was prophylaxed with iv immunoglobulin. Six weeks posttransplant, he presented with a febrile illness with abnormal liver enzymes, CMV antigenemia, and positive buffy coat cultures, which subsequently resolved without therapy. One month later, he developed a second rise in

1618

Peak

DayO

Figure 3. Serial CMV-AG disease.

count 0

of results of the CMV-AG coat

Ag

Control

CMV-AG level with no clinical evidence of disease. On Day 120, he was readmitted with febrile illness and pneumonitis and required ventilatory support. He was treated with gancicbovir and appeared to be responding when he suffered a fatal cardiac arrest. An autopsy revealed evidence of a fresh thrombosis of the left anterior descending coronary artery and residual CMV pneumonitis. This case illustrates the utility of serial monitoring of CMV-AG counts in foblowing the course of CMV infection and predicting clinical relapse.

DISCUSSION CMV infection continues to constitute a major cause of morbidity and occasional mortality in the transplant recipient. Particularly at risk is the seronegative recipient of an allograft from a seropositive donor; such patients often receive routine prophylaxis with immunoglobulin (10). high-dose acycbovir (1 1), or more recently, gancicbovir (1 2). Nevertheless,

Volume

4

Number

#{149}

8



1994

Murray

L.B.

Patnt: Blood Culture

+

-

-

+

-

-

O%O

0

et al

OO

2OO

0

n

1$

10

ii S

130

1o

Day Figure 4. Serial recipient

CMV-AG with self-limited

levels, alanine CMV disease.

aminotransferase

life-threatening infection can occur, and there is increasing evidence that even individuals seropositive before transplant may be at risk of reinfection or reactivation of disease, especially when anti-T cell immunotherapy is used (13). Effective antiviral therapy is now available with ganciclovir and other agents. Chemotherapy appears to be most efficacious when instituted early in the course of infection (14); however, early diagnosis of CMV infection has proved problematic. The detection of serologic evidence of infection (1gM’ significant rise in IgG) is dependent on an intact host humoral response, which may take several weeks to develop. The detection of vircmia, with positive buffy coat cultures, has often been correlated with the presence of significant infection, but conventional cell culture techniques, which depend on the detection of cytopathic effects on cubtured cells, take 2 to 3 wk. The shell vial technique

Journal

of the American

Society

of Nephrology

(ALT)

level,

and

white

blood

cell

count

in a renal

transplant

(1 5) was developed to achieve more rapid laboratory diagnosis by detecting CMV antigens in a fibroblast monolayer but still requires 24 to 72 h to obtain positive results. Van dcr Bij et at. (3) first reported the appearance of CMV-AG in the PBL of transplant recipients infected with CMV. In a subsequent publication, they suggested that this was a very sensitive and specific test for the diagnosis of CMV infection in renal transplant recipients that gives positive results very early in the condition (4). Our experience with the CMVAG assay appears to confirm the results of Van den Berg et at. (4)in that the CMV-Ag assay was sensitive, detecting all 1 8 patients with laboratory evidence of symptomatic CMV infection. Because a gold standard” test in the strictest sense is not available to confirm CMV infection and because both buffy coat culture and 1gM level were used in the laboratory,

1619

CMV-Antigenemia

Assay

RD.

Patient: Blood

Culture

+

+

+

pneumonitis V

fatal Ml V

Q200

oo/0\

100

,

0 200

150W .

0

,1OO, 0

‘0

50 OO

0. 20 0

E

.

15

I#{176}

.

0 0

U .

/#{176}%#{176}

I

10 0

I

I

30

#{149}0

I I 20

90

ISO

Day

Figure (See

5. Serial CMV-AG levels, alanine aminotransferase Text for details). MI, myocardial infarction.

(ALT) level,

confirmation of infection and a strict comparison of sensitivity and specificity with the CMV-Ag test could not be assessed. However, a significant advantage of the CMV-AG assay was the rapidity with which the result was available, invariably within 36 h, whereas the average time to a positive blood culture was 8 days, representing a delay of a week in establishing a diagnosis. The CMV-AG assay also appeared to be specific for CMV infection. Indeed, specificity would also have been 100% except for a single patient. This patient, who was seropositive pretransplant, developed unexplained fevers 6 wk posttransplant with no other manifestations of CMV infection. The initial CMVAG count was 3 celbs/50,000, with negative 1gM titers and buffy coat blood culture. The patient was subsequently admitted for a further workup of these

1620

and white

blood

cell count

in a CMV-infected

patient

fevers and was found to be leukopenic. A second CMV-AG count was 14 cells/50,000, but buffy coat cultures and 1gM titer remained negative. No cause for the fevers was identified, and they resolved spontaneousby. as did his leukopenia. Convalescent tests showed that CMV antigenemia disappeared. and buffy coat culture and 1gM titer remained negative. Although we feel that the patient most likely had CMV-rebated fever and leukopenia, it was not class!fied as such for the purposes of this study because of a lack of independent laboratory confirmation of the diagnosis. The CMV-AG assay was less specific in the diagnosis of CMV disease in that several patients with positive CMV-AG counts had fever as the only manifestation of infection (Figure 1 ; Table 2). The same was also true of buffy coat blood culture (Table 3).

Volume

4

Number

#{149}

8

1994

#{149}

Murray

However, one can take advantage of the quantitative nature of the CMV-AG test in that higher CMV-AG counts were associated with invasive disease. Indeed, if one used a cutoff of 10 cells/50,000 PBL, the positive predictive value for CMV disease rose to 93%. It should be emphasized that this applies to the initial CMV-AG assay drawn after the onset of fever. In contrast, when peak CMV-AG counts were analyzed, a value of >40 celbs/50,000 PBL was most predictive of CMV disease (Table 1). Not only was the CMV-AG test useful in the diagnosis of CMV disease, it also proved useful in monitoring the subsequent course of infection, with counts rising to a peak and subsequently falling as clinical manifestations resolved (Figures 3 to 5). Since the initial report of Van Den Berg et ab. (4), a number of other reports on the utility of the CMV antigen assay have appeared. Wunderbi et at. (8) reported in a prospective study that CMV-AG was less efficient than either CMV 1gM serology or buffy coat culture in the detection of CMV infection in transplants. Miller et at. (9) were unable to detect any antmgenemia in 27 transplant recipients tested prospectively. of whom 1 0 developed CMV disease, although the methodology used in this study has been questioned, in particular, the failure to use fresh specimens (16). Zipeto et at. (1 7) recently compared various modalities, includinga diagnostic pobymerase chain reaction assay, for detecting human CMV DNA. Their results are in agreement with those of Van Den Berg et at. (4) and ourselves, in that there appeared to be a quantitative correlation between antigenemia and viremia and overt clinical symptoms were gencralby observed when antigencmia exceeded >50 positive cells per 50,000 PBL. Interestingly, in that study, samples positive by polymerase chain reaction alone without antigenemia were never associated with clinical symptoms. Finally. both Bocckh et at. (6) and Erice et at. (7) have reported that the CMVAG assay was more sensitive than shell vial centrifugation cultures in the rapid detection of CMV infection. Boeckh et at. (6) also found that the bevel of antigenemia correlated with the severity of disease. Our approach was different from most of these studies in that CMV-AG assays were performed only after patients developed clinical manifestations of possible CMV infection and they were not monitored prospectively with regular CMV-AG determinations from the time of transplant. Others (4,6) have found that CMV antigenemia is often detectable even before clinical manifestations of CMV infection develop, raising the intriguing possibility of pharmacologic intervention (i.e. antiviral therapy) even before significant clinical manifestations appear. However, because of the cost and potential toxicity of such agents, further prospective studies will be required to see if such an approach can limit subsequent morbidity and mortality and at what threshold count ,

Journal

of the

American

Society

of Nephrology

et al

of the CMV-AG assay such intervention should be started. In summary, in renal transplant recipients with symptoms suggestive of CMV infection, a positive CMV-AG count was highly suggestive of symptomatic CMV infection and an initial titer of >10 cells 50,000 PBL was predictive of progression to significant CMV disease. In patients with CMV disease monitored serially, CMV-AG levels tended to rise to a peak and subsequently fall as clinical manifestations resolved. We concluded that the CMV-AG assay is a sensitive test in the early diagnosis of CMV infection in renal transplant recipients and provides more rapid results than existing diagnostic modalities (buffy coat cubture, anti-CMV 1gM). When expressed quantitatively, it can be useful in indicating the likelihood of invasive disease as well as in monitoring the course and severity of infection.

ACKNOWLEDGMENTS The authors thank Dr. T.H. The of the State University Hospital. Groningen, Netherlands for his help in establishing the antigenemia assay technique and for providing the necessary monoclonal antibodies.

REFERENCES 1 . Rubin RH: Impact of cytomegalovirus infection on organ transplant recipients. Rev Infect Dis 1 990;’l 2[Suppb ‘/I5754-5176. 2. Drew WL: Diagnosis of cytomcgalovirus infection. Rev Infect Dis 1988; 10:[Suppb 31:54685476. 3. Van der Blj W, Torensma R, Van Son WJ, et al.: Rapid immunodiagnosis of active cytomegalovirus infection by monocbonal antibody stainof blood leukcocytes. J Med Virol 1988;25: 179-188. 4. Van den Berg AP, Van der Bij W, Van Son WJ, et at.: Cytomegalovirus antigenemia as a useful marker of symptomatic cytomegalovirus infection after renal transplantation-a report of 1 30 consecutive patients. Transplantation 1989; 48:991-995. 5. Revello MG, Percivalle E, Zavattoni M, Parea M, Grossi P, Gerna G: Detection of human cytomegalovirus immediate early antigen in leucocytes as a marker of vircmia in immunocompromised patients. J Med Virol 1989;29:88-93. 6. Boeckh M, Bowden RA, Goocirish JM, Pettinger M, Meyers JD: Cytomegalovirus antigen detection in peripheral bboocFbcukocytcs after albogeneic marrow transplantation. Blood 1992;80: 1358-1364. 7. Erice A, Hoim MA, Gill PC, et al.: Cytomegalovirus (CMV) antigenemia assay is more sensitive than shell viaF cultures for rapid detection of CMV in polymorphonucbear blood leukocytes. J Clin Microbiob 1992;30:2822-2825. 8. Wunderli W, Auracher JSD, Zbinden R: Cytomegabovirus detection in transplant patients: Comparison of different methods in a prospective survey. J Clin Microbiob 1991 ;29:26482650.

mug

1621

Murray

9.

1 0.

1 1.

12.

1 3.

1622

Miller HE, Rossler E, Milk R, Thomas C: Prospective study of cytomegabovirus antigenemia in albograft recipients. J Clin Microbiol 1991;29:1054-1055. Snydman DR, Werner BG, Heinze-Lacey B, et at.: Use of cytomegabovirus immune globulin to prevent cytomegabovirus disease in renal transplant recipients. N Engl J Med 1987;317:1049I 054. Balfour 1111, Cbace BA, Stapleton JT, Simmons RL, Fryd DS: A randomized, placebo-controlled trial of orab acyclovir for the prevention of cytomegalovirus disease in recipients of renal albograrts. N Engl J Med 1989;320:1381-1387. chmldt GM, Horak DA, Niland JC, et a!.: A randomized controlled trial of prophhylactic gancicbovir for cytomegalovirus pulmonary infection in recipients of aibogeneic bone marrow transplants. N Engl J Med 1991;324:10551011. Hibberd PL, Tolkoff-Rubin NE, Cosimi AB, et al.: Symptomatic cytomegalovirus disease in the

14.

1 5.

16. 1 7.

et aI

cytomegalovirus antibody seropositive renal transplant recipient treated with OKT3. Transplantation 1 992;53:68-72. Snydman DR: Ganciclovir therapy for cytomegabovirus disease associated with renal transplants. Rev Infect Dis 1988;lOISuppl 3J:5554562. Paya C, Smith TF, Ludwig J, Hermans PE: Rapid shell vial culture and tissue histology compared with serology for the rapid diagnosis of cytomegabovirus infection in liver transplantation. Mayo Clin Proc 1988;64:670-675. Van der Glessen M, The TH, Van Son WJ: Cytomegalovirus antigenemia assay. J Clin Microbiob 191 ;29:2909-2910. Zlpeto D, Revello MG, Silini E, Parea M, Percivalle E: Development and clinical significance of a diagnostic assay based on the pobymerase chain reaction for detection of human cytomegabovirus DNA in blood samples from immunocompromised patients. J Clin Microbiol 1992; 30:527-530.

Volume

4

Number

#{149}

8

.

1994