THE JOURNAL OF INFECTIOUS DISEASES· VOL. 132, NO.4· OCTOBER 1975 © 1975 by the University of Chicago. All rights reserved.
Epidemiology of Cytomegalovirus Infection after Transplantation and Immunosuppression
Viral infections and clinical complications were studied during hemodialysis and after renal transplantation. Active cytomegalovirus infection developed in 96% of patients after renal transplantation; reactivation of herpes simplex, varicella-zoster, and Epstein-Barr viruses was found in 35%,24%, and 0% of patients, respectively. Cytomegalovirus viremia developed in 42% of patients an average of two months after renal transplantation, lasted 1.75 (± 1.5) months (except in one patient with chronic viremia), and was followed by chronic viruria. Higher titers of infectious cytomegalovirus were found in the polymorphonuclear than in the mononuclear leukocyte fraction. Reactivation of a latent infection and, less likely, respiratory infection appear to be the most probable mechanisms of cytomegalovirus infection after renal transplantation. One to three months after transplant, cytomegalovirus infection may be related to fever, arthralgia, pneumonitis, and leukopenia; three to four months after transplant, the virus may be related to hepatitis; and 12-30 months after transplant, it may be related to retinitis in patients with chronic viremia. Although other causes of these complications are possible, herpes simplex virus, Epstein-Barr virus, varicella-zoster virus, measles virus, adenovirus, hepatitis B virus, and Toxoplasma gondii appear to be of lesser importance than cytomegalovirus in this respect.
Active cytomegalovirus (CMV) infection occurs frequently in renal transplant recipients [1, 2]. Rates of CMV infection, based on significant CF antibody responses and on the presence of vi-
ruria, have been found to be as high as 9 99% purity (except for red cells), as determined by examination of 100cells. Red blood cells were obtained from the bottom of the red cell pellet in the Ficoll- Hypaque gradient. Titers of virus in mononuclear and polymorphonuclear leukocytes were determined on the basis of the number (ni) of virus foci (foci ofswollen or rounded refractile cells that increased in size on serial observations and, after staining,
Downloaded from http://jid.oxfordjournals.org/ at University of York on August 20, 2014
Patients and Methods
Fiala et al.
423
CMV Infection after Transplantation
displayed characteristic intranuclear inclusions) one week after the first detection of virus C PE. Only counts of < 15 foci per well were used in the calculation. Virus titer was determined as: (n 1
+ n' 1 + n2 + n '2 + + n's) 2N 1 + 2N2 + 2Ns
X
105
I
1 J. Murnane, M. Fiala, R. Wallace, L. B. Guze, and D. C. Heiner, "Proteins of Cytomegalovirus. II. Constituents of Soluble Cytomegalovirus Complement-Fixing Antigen," manuscript in preparation.
Downloaded from http://jid.oxfordjournals.org/ at University of York on August 20, 2014
focus-forming units (FFU)/cell, where n i and n i are duplicate counts in wells inoculated with N i cells. Ninety-five percent confidence interval of the titer was found using a table [14]. Virus was identified as CMV on the basis of its slow growth in human fibroblast cultures with production of characteristic intranuclear inclusions and, on electron microscopy, enveloped and naked particles with typical herpesvirus morphology and "dense bodies" [15]. CF studies with antigens of CMV (A D 169 strain), herpes simplex virus (HSV), varicellazoster virus (VZV), measles virus, and adenovirus were performed by the microtiter technique modification of the Kolmer technique (California State Health Department) using 2 exact units of antigen, 2 units of hemolysin, and 2 exact units of complement. The highest dilution of serum resulting in 3+ or 4+ fixation was considered as its C F titer. Hepatitis B surface antigen (HB s Ag) and antibody (anti- HB J were determined by a direct solid-phase radioimmunoassay technique [16]. Specimens positive for anti-Hls, were confirmed by neutralization with human serum containing anti- HB s • Immunofluorescent antibodies to the capsid antigen and early antigen of Epstein-Barr virus (EB V) were determined as described previously [17, 18]. Immunofluorescent antibody to Toxoplasma gondii was determined by an indirect technique using slides with T. gondii organisms (Electronucleonics, Bethesda, Md.) and rabbit antiserum to human globulin (Beckmann, Palo Alto, Calif.). In all patients, clinical history (in particular, fever, arthralgia, cough, rashes, jaundice, and chest and back pain) and laboratory data (complete blood count, transaminase enzyme levels, bilirubin, serum and urine electrolytes, uric acid and creatinine, and urine culture for bacteria) were obtained regularly. Blood, urine, sputum, and throat were cultured for bacteria and fungi and, in the cases of blood, urine, and some
liver biopsy specimens, also for viruses. Cytologic tests of urine were performed with the Papanicolau stain. Bronchoscopy was performed when indicated by the severity of pulmonary disease. Sputum was then obtained by bronchial washing for bacterial and fungal cultures. Bronchial brush biopsy and/or alveolar biopsy and sputum cytology were also done when indicated. Specimens were stained with silver methenamine sulfate and cresyl echt violet [19] for detection of Pneumocystis carinii. Patients with elevations of serum glutamicoxalacetic transaminase (SGOT) or serum glutamic-pyruvic transaminase (SGPT) above 200 Karmen units/rnl, or with elevation of the bilirubin above 2 mg/IOO ml were usually changed from azathioprine to cyclophosphamide therapy [20]. Other hepatotoxic drugs were discontinued if possible. Extrahepatic biliary obstruction was excluded by appropriate radiological techniques. Percutaneous liver biopsy was performed in 13 patients. Renal allograft rejection was diagnosed on clinical and laboratory grounds. The laboratory criteria included a sustained rise of serum creatinine of at least 0.3 mg/ 100 ml and a fall in creatinine clearance and urinary excretion of sodium. In some instances, other causes of impaired renal function were excluded by renal scintigrams, iv pyelograrns , or, when necessary, biopsy of the allograft. Immunofluorescence studies and electron microscopy were performed on all renal allograft biopsy specimens. Immunofluorescence studies for detection of fibrinogen, f3-1 C globulin, Ig G, IgA, IgM, IgE, and CMV antigen were performed. Human IgG antibody to CMV prepared from a specimen of serum from a patient convalescent from C MV mononucleosis was conjugated with fluorescein isothiocyanate (FITC) according to the method of Clark and Shephard [21]. Hyperimmune rabbit antiserum to isolated soluble C MV C F antigen 1 was used by the indirect immunofluorescence technique. This serum
Fiala et al.
424
Results
Epidemiology (studies A and B). The prevalence of CMV viremia in studies A and B was significantly higher in patients observed during the first year after transplantation (23 of 43 patients, or 53%) than in those observed during the second year (three of 19, or 16%) (P < 0.005). Rates of CMV viruria were similar in the two years (28 of 41, or 68%, and nine of 19, or 47%, respectively; P > 0.05). Patients studied during the first year after transplantation had higher CF titers to CMV than those studied during the sec-
ond year (P = 0.007) (table 1). CF antibody to CMV frequently declined to low or undetectable levels after clearing of the viremia. The incidence of active cases of CMV infection in study A (determined by isolation of virus from blood or urine and/or by a significant rise in titer of CF antibody) was zero of 19 during the period before transplantation, nine of26 (35%)during the first month, nine of 26 (35%) during the second month, two of26 (8%)during the third month, one of26 (4%) during the fourth month, two of26 (8%) during the fifth month, one of 26 (4%) during the sixth month, and one of 26 (4%) during the seventh month after transplantation. Infectious virus in the blood of viremic patients was cell-associated, since plasma or plasma pellets never yielded virus, whereas repeated washing of leukocytes before inoculation into cell culture did not decrease titers of virus. Infectious virus was found only in polymorphonuclear and mononuclear cell fractions. Titers of virus were 2.3-7.7 times greater in polymorphonuclear than in mononuclear cells when both fractions were positive, and in some instances virus was found only in the polymorphonuclear fraction (table 2). Interestingly, a patient with CMV mononucleosis whose viremia was investigated by the same technique had a similar predominance of infectious virus in polymorphonuclear cells (authors' unpublished observations). Only 14 of 26 patients who had undergone transplantation received blood during or after the transplant surgery. Active CMV infection developed in all 14 patients who received blood in the perioperative period; CMV infection developed in 11 of the remaining 12 patients. Relation of clinicalfindings to active CMV infection. Prolonged fever, transient arthralgia, or arthralgia with fever occurred in 20 (77%)of the 26
Table 1. Prevalence of cytomegalovirus (CMV) viremia and viruria and CF antibody in relation to the interval after transplantation (studies A and B combined). Highest reciprocal CMV CF antibody titer
C M V isolation Interval* 1-12 13-24 ~25
Viremia
Viruria
8 mg/l00 ml) in two patients. The level of calcium in serum was < 10.5 mg/l00 ml in
Downloaded from http://jid.oxfordjournals.org/ at University of York on August 20, 2014
5
Intervalt
Blood cell fractions
Fiala et al,
426
Table 3. Onsets of abnormal virological, clinical, and laboratory findings after renal transplantation and maximal titers of CF antibody to cytomegalovirus (CMV) (study A). Clinical and laboratory findings* CMV isolation* Titer of CF antibody Patient - - - - - - - - - - - - - no. Blood Urine Pre t Highestf Riser I
-II
2 3
137 55 38 14 59 15
4 5 6 7 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 Total positive
42
47
25 105
158 11/26 (42.3)
1:128 1:32 1:128 1:256 1:64 1:256 1:32 1:64 1:32 1:64 1:128 1:128 1:1,024
Epidemiology of Cytomegalovirus Infection after Transplantation and Immunosuppression
Viral infections and clinical complications were studied during hemodialysis and after renal transplantation. Active cytomegalovirus infection developed in 96% of patients after renal transplantation; reactivation of herpes simplex, varicella-zoster, and Epstein-Barr viruses was found in 35%,24%, and 0% of patients, respectively. Cytomegalovirus viremia developed in 42% of patients an average of two months after renal transplantation, lasted 1.75 (± 1.5) months (except in one patient with chronic viremia), and was followed by chronic viruria. Higher titers of infectious cytomegalovirus were found in the polymorphonuclear than in the mononuclear leukocyte fraction. Reactivation of a latent infection and, less likely, respiratory infection appear to be the most probable mechanisms of cytomegalovirus infection after renal transplantation. One to three months after transplant, cytomegalovirus infection may be related to fever, arthralgia, pneumonitis, and leukopenia; three to four months after transplant, the virus may be related to hepatitis; and 12-30 months after transplant, it may be related to retinitis in patients with chronic viremia. Although other causes of these complications are possible, herpes simplex virus, Epstein-Barr virus, varicella-zoster virus, measles virus, adenovirus, hepatitis B virus, and Toxoplasma gondii appear to be of lesser importance than cytomegalovirus in this respect.
Active cytomegalovirus (CMV) infection occurs frequently in renal transplant recipients [1, 2]. Rates of CMV infection, based on significant CF antibody responses and on the presence of vi-
ruria, have been found to be as high as 9 99% purity (except for red cells), as determined by examination of 100cells. Red blood cells were obtained from the bottom of the red cell pellet in the Ficoll- Hypaque gradient. Titers of virus in mononuclear and polymorphonuclear leukocytes were determined on the basis of the number (ni) of virus foci (foci ofswollen or rounded refractile cells that increased in size on serial observations and, after staining,
Downloaded from http://jid.oxfordjournals.org/ at University of York on August 20, 2014
Patients and Methods
Fiala et al.
423
CMV Infection after Transplantation
displayed characteristic intranuclear inclusions) one week after the first detection of virus C PE. Only counts of < 15 foci per well were used in the calculation. Virus titer was determined as: (n 1
+ n' 1 + n2 + n '2 + + n's) 2N 1 + 2N2 + 2Ns
X
105
I
1 J. Murnane, M. Fiala, R. Wallace, L. B. Guze, and D. C. Heiner, "Proteins of Cytomegalovirus. II. Constituents of Soluble Cytomegalovirus Complement-Fixing Antigen," manuscript in preparation.
Downloaded from http://jid.oxfordjournals.org/ at University of York on August 20, 2014
focus-forming units (FFU)/cell, where n i and n i are duplicate counts in wells inoculated with N i cells. Ninety-five percent confidence interval of the titer was found using a table [14]. Virus was identified as CMV on the basis of its slow growth in human fibroblast cultures with production of characteristic intranuclear inclusions and, on electron microscopy, enveloped and naked particles with typical herpesvirus morphology and "dense bodies" [15]. CF studies with antigens of CMV (A D 169 strain), herpes simplex virus (HSV), varicellazoster virus (VZV), measles virus, and adenovirus were performed by the microtiter technique modification of the Kolmer technique (California State Health Department) using 2 exact units of antigen, 2 units of hemolysin, and 2 exact units of complement. The highest dilution of serum resulting in 3+ or 4+ fixation was considered as its C F titer. Hepatitis B surface antigen (HB s Ag) and antibody (anti- HB J were determined by a direct solid-phase radioimmunoassay technique [16]. Specimens positive for anti-Hls, were confirmed by neutralization with human serum containing anti- HB s • Immunofluorescent antibodies to the capsid antigen and early antigen of Epstein-Barr virus (EB V) were determined as described previously [17, 18]. Immunofluorescent antibody to Toxoplasma gondii was determined by an indirect technique using slides with T. gondii organisms (Electronucleonics, Bethesda, Md.) and rabbit antiserum to human globulin (Beckmann, Palo Alto, Calif.). In all patients, clinical history (in particular, fever, arthralgia, cough, rashes, jaundice, and chest and back pain) and laboratory data (complete blood count, transaminase enzyme levels, bilirubin, serum and urine electrolytes, uric acid and creatinine, and urine culture for bacteria) were obtained regularly. Blood, urine, sputum, and throat were cultured for bacteria and fungi and, in the cases of blood, urine, and some
liver biopsy specimens, also for viruses. Cytologic tests of urine were performed with the Papanicolau stain. Bronchoscopy was performed when indicated by the severity of pulmonary disease. Sputum was then obtained by bronchial washing for bacterial and fungal cultures. Bronchial brush biopsy and/or alveolar biopsy and sputum cytology were also done when indicated. Specimens were stained with silver methenamine sulfate and cresyl echt violet [19] for detection of Pneumocystis carinii. Patients with elevations of serum glutamicoxalacetic transaminase (SGOT) or serum glutamic-pyruvic transaminase (SGPT) above 200 Karmen units/rnl, or with elevation of the bilirubin above 2 mg/IOO ml were usually changed from azathioprine to cyclophosphamide therapy [20]. Other hepatotoxic drugs were discontinued if possible. Extrahepatic biliary obstruction was excluded by appropriate radiological techniques. Percutaneous liver biopsy was performed in 13 patients. Renal allograft rejection was diagnosed on clinical and laboratory grounds. The laboratory criteria included a sustained rise of serum creatinine of at least 0.3 mg/ 100 ml and a fall in creatinine clearance and urinary excretion of sodium. In some instances, other causes of impaired renal function were excluded by renal scintigrams, iv pyelograrns , or, when necessary, biopsy of the allograft. Immunofluorescence studies and electron microscopy were performed on all renal allograft biopsy specimens. Immunofluorescence studies for detection of fibrinogen, f3-1 C globulin, Ig G, IgA, IgM, IgE, and CMV antigen were performed. Human IgG antibody to CMV prepared from a specimen of serum from a patient convalescent from C MV mononucleosis was conjugated with fluorescein isothiocyanate (FITC) according to the method of Clark and Shephard [21]. Hyperimmune rabbit antiserum to isolated soluble C MV C F antigen 1 was used by the indirect immunofluorescence technique. This serum
Fiala et al.
424
Results
Epidemiology (studies A and B). The prevalence of CMV viremia in studies A and B was significantly higher in patients observed during the first year after transplantation (23 of 43 patients, or 53%) than in those observed during the second year (three of 19, or 16%) (P < 0.005). Rates of CMV viruria were similar in the two years (28 of 41, or 68%, and nine of 19, or 47%, respectively; P > 0.05). Patients studied during the first year after transplantation had higher CF titers to CMV than those studied during the sec-
ond year (P = 0.007) (table 1). CF antibody to CMV frequently declined to low or undetectable levels after clearing of the viremia. The incidence of active cases of CMV infection in study A (determined by isolation of virus from blood or urine and/or by a significant rise in titer of CF antibody) was zero of 19 during the period before transplantation, nine of26 (35%)during the first month, nine of 26 (35%) during the second month, two of26 (8%)during the third month, one of26 (4%) during the fourth month, two of26 (8%) during the fifth month, one of 26 (4%) during the sixth month, and one of 26 (4%) during the seventh month after transplantation. Infectious virus in the blood of viremic patients was cell-associated, since plasma or plasma pellets never yielded virus, whereas repeated washing of leukocytes before inoculation into cell culture did not decrease titers of virus. Infectious virus was found only in polymorphonuclear and mononuclear cell fractions. Titers of virus were 2.3-7.7 times greater in polymorphonuclear than in mononuclear cells when both fractions were positive, and in some instances virus was found only in the polymorphonuclear fraction (table 2). Interestingly, a patient with CMV mononucleosis whose viremia was investigated by the same technique had a similar predominance of infectious virus in polymorphonuclear cells (authors' unpublished observations). Only 14 of 26 patients who had undergone transplantation received blood during or after the transplant surgery. Active CMV infection developed in all 14 patients who received blood in the perioperative period; CMV infection developed in 11 of the remaining 12 patients. Relation of clinicalfindings to active CMV infection. Prolonged fever, transient arthralgia, or arthralgia with fever occurred in 20 (77%)of the 26
Table 1. Prevalence of cytomegalovirus (CMV) viremia and viruria and CF antibody in relation to the interval after transplantation (studies A and B combined). Highest reciprocal CMV CF antibody titer
C M V isolation Interval* 1-12 13-24 ~25
Viremia
Viruria
8 mg/l00 ml) in two patients. The level of calcium in serum was < 10.5 mg/l00 ml in
Downloaded from http://jid.oxfordjournals.org/ at University of York on August 20, 2014
5
Intervalt
Blood cell fractions
Fiala et al,
426
Table 3. Onsets of abnormal virological, clinical, and laboratory findings after renal transplantation and maximal titers of CF antibody to cytomegalovirus (CMV) (study A). Clinical and laboratory findings* CMV isolation* Titer of CF antibody Patient - - - - - - - - - - - - - no. Blood Urine Pre t Highestf Riser I
-II
2 3
137 55 38 14 59 15
4 5 6 7 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 Total positive
42
47
25 105
158 11/26 (42.3)
1:128 1:32 1:128 1:256 1:64 1:256 1:32 1:64 1:32 1:64 1:128 1:128 1:1,024
