Molecular Methods for Cytomegalovirus Surveillance in Bone Marrow

0 downloads 0 Views 65KB Size Report
Adriana Weinberg,1,2* Debra Schissel,3 and Roger Giller1,3. Departments of ... compared by using plasma, peripheral blood leukocytes (PBLs), and whole blood (WB) from allogeneic bone .... 80°C. On the day of the assay, DNA was.
JOURNAL OF CLINICAL MICROBIOLOGY, Nov. 2002, p. 4203–4206 0095-1137/02/$04.00⫹0 DOI: 10.1128/JCM.40.11.4203–4206.2002 Copyright © 2002, American Society for Microbiology. All Rights Reserved.

Vol. 40, No. 11

Molecular Methods for Cytomegalovirus Surveillance in Bone Marrow Transplant Recipients Adriana Weinberg,1,2* Debra Schissel,3 and Roger Giller1,3 Departments of Pediatrics1 and Medicine,2 University of Colorado, and The Children’s Hospital,3 Denver, Colorado Received 29 May 2002/Returned for modification 29 July 2002/Accepted 26 August 2002

Two different methods for detection of cytomegalovirus (CMV), PCR and hybrid capture (HC), were compared by using plasma, peripheral blood leukocytes (PBLs), and whole blood (WB) from allogeneic bone marrow transplant recipients. One hundred specimens were obtained from nine children over an 18-month surveillance period. PCR of plasma for CMV was used for clinical management. The proportions of samples positive for CMV DNA by PCR with plasma, HC with WB, and PCR with PBLs were 21, 28, and 37%, respectively. Among 44 samples that were tested by all three methods, 68% had concordant results. By using a robust definition of true-positive samples (positivity by two or more methods or positivity of sequential samples by one method), the sensitivities of PCR with plasma, HC with WB, and PCR with PBLs were 50, 67, and 83%, respectively, and the specificities were 100, 96, and 96%, respectively. Two patients developed CMV-associated end-organ disease (one developed respiratory disease, and one developed gastrointestinal disease). CMV DNA was not detected in the plasma 1 week prior to the development of symptoms in either patient, whereas HC with WB was positive for both patients and PCR with PBLs was for one patient. These data suggest that WB or PBLs might be the preferred sample for use for surveillance for CMV in immunocompromised patients. decreasing the rate of mortality among BMT recipients (3, 11, 22). Treatment protocols based on detection of CMV antigenemia or molecular detection of circulating CMV DNA are also successful and are also simpler in preventing CMV disease in BMT and solid-organ transplant patients (1, 2, 13, 24, 27). The objective of this study was to compare two different methodologies, PCR and hybrid capture (HC), for the detection of CMV DNA in different blood compartments: plasma, peripheral blood leukocytes (PBLs), and whole blood (WB). It also provides the first direct comparison of two semiautomated platforms that are commercially available for detection of CMV DNA. (Part of this study was presented at the 18th Clinical Virology Symposium, Clearwater, Fla., April 2002.)

Cytomegalovirus (CMV) infection is the most common and potentially devastating viral opportunistic infection in bone marrow transplant (BMT) recipients (6, 16, 28). Prior to the introduction of prophylactic and preemptive antiviral treatment strategies, approximately 50% of BMT recipients with active CMV infection developed end-organ disease such as pneumonitis, gastroenteritis, and, more rarely, hepatitis, retinitis, and neurologic disorders. The case fatality rate of CMV pneumonitis in these patients remains very high (between 30 and 70%), despite the introduction of CMV-specific antiviral therapy (10). Two main strategies are used to prevent CMV disease in BMT recipients: antiviral anti-CMV prophylaxis of all CMVseropositive patients during the first 100 days after transplantation, which is the interval of peak incidence of CMV disease, or preemptive antiviral therapy, whereby treatment is initiated upon detection of asymptomatic CMV reactivation in blood. Prophylactic therapy, although very successful in decreasing the incidence of CMV disease, does not affect the overall rate of mortality among BMT recipients (12). The success of the alternative strategy, preemptive therapy, depends on the sensitivity of the detection methods and the adequacy of the clinical specimen tested. Previous studies have demonstrated that culture of blood, urine, or nasopharyngeal secretions for CMV has low positive and negative predictive values with respect to CMV end-organ disease (17). In contrast, preemptive therapy guided by culture of bronchoalveolar lavage specimens for CMV is highly successful in preventing CMV pneumonia and

MATERIALS AND METHODS Patients and study design. This study prospectively enrolled CMV-seropositive pediatric BMT patients receiving CMV surveillance according to the current standard of care between May 2000 and August 2001. Plasma from the patients was tested for CMV by PCR at 7 to 10 days pretransplantation and then weekly posttransplantation from day ⫹14 to day ⫹120. PCR testing was continued for a longer period for patients still on immunosuppressive therapy at day ⫹120. Patients in whom CMV DNA was detected in plasma by PCR received therapeutic doses of ganciclovir (5 mg/kg of body weight every 12 h) for at least 10 days or until the result of the plasma PCR test reverted to negative. Thereafter, maintenance doses of ganciclovir (5 mg/kg five times per week) were given for 4 weeks or until the patient’s immunosuppressive therapy was reduced to minimal levels. Qualitative PCR for CMV in plasma. A qualitative PCR for CMV in plasma was performed by a previously published technique with primers 625 and 461 and probe 626 from the EcoRI fragment D region of CMV strain AD 169 (25, 26). DNA was extracted from 200 ␮l of plasma with QIAamp DNA purification columns (Qiagen, Valencia, Calif.). The amplification was carried out in a 50-␮l final volume containing 20 ␮l of extracted DNA, 0.5 U of PfuI (Stratagene, La Jolla, Calif.), each of the four deoxynucleoside triphosphates at a concentration of 100 ␮M, and 1 ␮M each primer in PfuI buffer (Stratagene). The amplified DNA was separated by agarose gel electrophoresis and transferred to nylon

* Corresponding author. Mailing address: University of Colorado Health Sciences Center, 4200 East 9th Ave., Campus Box C227, Denver, CO 80262. Phone: (303) 315-4624. Fax: (303) 315-1787. E-mail: [email protected]. 4203

4204

WEINBERG ET AL.

membranes. The identities of the DNA bands were confirmed by hybridization with digoxigenin-labeled probe 626 (Boehringer Mannheim Biochemicals, Indianapolis, Ind.), followed by treatment with an antidigoxigenin antibody conjugated to alkaline phosphatase (Boehringer Mannheim) and the chemiluminescent substrate for phosphatases (Tropix, Bedford, Mass.). Each assay included a negative control consisting of water and two positive sensitivity controls consisting of a detergent lysate of CMV AD 169 at 0.01 PFU/reaction tube and a clinical specimen known to be positive. The test was considered valid if the controls yielded the expected results and the results of replicate tests with patient samples agreed. This method has an analytical sensitivity of 100 copies/ml. Quantitative HC for CMV. Quantitative HC for CMV was performed quantitatively with Hybrid Capture CMV DNA assay (version 2.0) kits (Digene). The assay consisted of RNA denaturation, hybridization with multiple CMV-specific RNA probes, and detection of the DNA-RNA hybrids with hybrid-specific monoclonal antibodies and a chemoluminescent substrate. The number of CMV DNA copies per milliliter of sample was calculated by manufacturer-provided software that interpolated the chemiluminescence readings of each sample on a linear regression curve generated with CMV DNA standards. The sensitivity of this test is ⱖ300 copies/ml, with a dynamic range from 600 to 570,000 copies/ml. For the purpose of the statistical analysis, an arbitrary value of 300 copies/ml was ascribed to samples containing 300 to 600 DNA copies/ml. Quantitative PCR for CMV in PBLs. The COBAS AMPLICOR CMV MONITOR kit (Roche Molecular Systems) adapted for PBLs was used for quantitative PCR for CMV in PBLs. This is a semiautomated competitive quantitative PCR that targets CMV DNA polymerase gene UL54 (2). The assay used 500 ␮l of EDTA-anticoagulated blood. After red blood cell lysis, leukocytes were washed, pelleted, and stored at ⫺80°C. On the day of the assay, DNA was extracted and added to the PCR master mixture containing the internal quantitative standard. DNA was amplified in the COBAS automated platform. Results generated by the CMV-specific COBAS software were adjusted for the PBL procedure and were expressed as the number of DNA copies per milliliter of WB. The test is quantitative for CMV loads ⱖ400 DNA copies/ml of plasma or ⱖ160 DNA copies/ml of WB, although the detection limit is ⱖ100 copies/ml of plasma or ⱖ40 copies/ml of WB. Samples containing 40 to 160 CMV DNA copies/ml of WB were ascribed an arbitrary value of 40 copies/ml for analysis. Statistical analysis. Statistical analysis was performed with Statview (version 5.0.1) software (SAS). P values ⱕ0.05 were considered significant.

RESULTS Patient characteristics. Nine CMV-seropositive allogeneic transplant recipients between 6 and 21 years of age, including five females and four males, were enrolled in the study. All patients had hematologic malignancies as their underlying disorder. Eight patients received cord blood cells from unrelated donors, and one received a bone marrow transplant from a matched sibling. Patients entered the study at a mean ⫾ standard deviation of 70 ⫾ 44 days after transplantation (range, ⫺5 to 162 days), and each patient had 2 to 33 blood draws over the 18-month period of observation. Patients received preemptive therapy with ganciclovir or foscarnet on seven occasions, but on two occasions, the CMV reactivation was not detected and the patients developed CMV end-organ disease. One patient had pneumonitis diagnosed by the detection of acute respiratory symptoms accompanied by a chest radiographic examination consistent with pneumonitis and simultaneous isolation of CMV from bronchoalveolar lavage specimens. A second patient had persistent bloody diarrhea in association with the isolation of CMV from blood and the identification of CMV in a tissue biopsy specimen by histochemistry. Both patients recovered from their illnesses after prolonged antiviral therapy. Comparative analysis of PCR with plasma, HC with WB, and PCR with PBLs for detection of CMV. Among the 100 blood samples collected in this study, 95, 95, and 51 samples were tested by PCR with plasma, HC with WB, and PCR with PBLs, respectively. The proportions of positive results gener-

J. CLIN. MICROBIOL. TABLE 1. Comparative results of molecular assays for detection of CMV DNA in different blood components No. of positive samples with the indicated result by the following assay: Assay and result

PCR with plasma Positive Negative

PCR with WB Positive

Negative

Positive

Negative

17 (19)a 7 (8)

3 (3) 63 (70)

8 (17) 8 (17)

1 (2) 29 (64)

12 (25) 7 (14)

4 (8) 26 (53)

HC with WB Positive Negative a

PCR with PBLs

Values in parentheses are percentages.

ated by PCR with plasma, HC with WB, and PCR with PBLs were 21, 28, and 37%, respectively. The rates of concordance of the results of PCR with plasma and HC with WB, PCR with plasma and PCR with PBLs, and HC with WB and PCR with PBLs were 89, 81, and 78%, respectively (Table 1). The relative sensitivities of PCR with plasma versus HC with WB and PCR with plasma versus PCR with PBLs were 74 and 53%, respectively; the sensitivities of HC with WB versus HC with plasma and PCR with PBLs were 89 and 70%, respectively; the sensitivities of PCR with PBLs versus PCR with plasma and PCR with PBLs versus HC with WB were 94 and 83%, respectively. PCR with PBLs generated a significantly higher number of positive results compared with the number generated by PCR with plasma (P ⫽ 0.007 by the chi-square test). The differences in the proportions of positive results generated by HC with WB versus PCR with PBLs or PCR with plasma versus HC with WB did not reach statistical significance (P ⫽ 0.5 and 0.2, respectively). The analysis of the results for the 44 specimens tested by all three methods showed a rate of concordance of 68% among all methods. We defined true-positive samples as those in which CMV DNA was detected by two or more methods or by one method if the positive result was preceded or followed in ⱕ1 week by other positive test results for CMV DNA. By this approach, the sensitivities of PCR with plasma, HC with WB, and PCR with PBLs were 50, 67, and 83%, respectively, and the specificities were 100, 96, and 96%, respectively. Differences in sensitivities and specificities did not reach statistical significance (P ⬍ 0.1 by chi-square analysis). There was a significant correlation (P ⬍ 0.0001 by linear regression) between the results of the two quantitative tests: HC with WB and PCR with PBLs (Fig. 1). Furthermore, paired analysis of these results did not show significant differences in the number of copies measured by the two methods (P ⫽ 0.2 by the paired t test). Clinical utility of PCR with plasma, HC with WB, and PCR with PBLs. PCR with plasma was performed in real time and was used to guide clinical management. On the basis of its results, seven patients received preemptive therapy against CMV. CMV reactivation was not detected by PCR with plasma in two patients, which allowed the progression of the infection to symptomatic disease at 56 and 170 days posttransplantation. In contrast, blood samples collected from these two patients 1 week before the initiation of symptoms were positive by HC

VOL. 40, 2002

MOLECULAR METHODS FOR SURVEILLANCE FOR CMV

FIG. 1. Correlation between circulating CMV loads in WB and in PBLs measured by HC and PCR, respectively. The data, which represent the log10-transformed number of CMV DNA copies per milliliter of blood, were derived from the results for 49 blood samples from nine pediatric allogeneic BMT patients. The equation describes the relationship between the measurements obtained by HC and PCR. The data are uniformly distributed on both sides of the diagonal, representing the line of equivalency, which indicates that there was no significant difference between measurements. The correlation between the two assays was highly significant, with P ⬍ 0.0001 (linear regression test).

with WB, but the blood sample from one patient was positive and the blood sample from the other patient was negative by PCR with PBLs. This suggests that HC with WB and PCR with PBLs might have superior sensitivities and negative predictive values compared with those of PCR with plasma. DISCUSSION The data reported here indicate that molecular detection of CMV DNA is more sensitive when WB is used than when plasma is used, which is in agreement with previous reports (9, 19, 30). The relative sensitivities of PCR with plasma, PCR with PBLs, and HC with WB were 50, 67, and 83%, respectively, when the results for specimens tested by all three methods were analyzed. The differences did not reach statistical significance due to the limited number of specimens (n ⫽ 44) tested by all three methods. However, the analysis of the results for the 46 specimens tested by HC with WB and PCR with plasma showed a significantly higher rate of detection of CMV DNA in WB than in plasma (P ⫽ 0.007). The specificities of PCR with plasma, HC with WB, and PCR with PBLs were 100, 96, and 96%, respectively. Furthermore, there was a rate of high concordance (68%) among the results of the three assays evaluated in this study, attesting to the high degree of accuracy of positive results generated by any of the three methods used in this study.

4205

This study provides the first direct comparison between the COBAS AMPLICOR CMV MONITOR assay with WB and the Hybrid Capture CMV DNA assay. Previous studies showed that the COBAS AMPLICOR CMV MONITOR assay with plasma was significantly less sensitive than HC with WB (A. Erice, C. Tierney, M. Hirsch, A. Caliendo, A. Weinberg, M. Kendall, and B. Polsky for the ACTG 360 Study Team, Abstr. 9th Conf. Retrovir. Opportunistic Infect., abstr. 38, 2002). This difference disappears when both assays use the same type of specimen, WB. Furthermore, we found a significant association between the numbers of CMV DNA copies per milliliter measured by these assays (P ⬍ 0.0001). The ability to quantitate CMV DNA can be used to improve the ability to predict the occurrence of CMV disease and avoid unnecessary antiviral therapy. Preemptive therapy might be more precisely selected by using the kinetics or threshold of the CMV load (8, 21, 29). In addition, quantitative measurements of CMV DNA can be used to monitor responses to therapy, since it has been demonstrated that the CMV load falls rapidly in the presence of effective therapy (27). With respect to complexity, cost, and turnaround time, HC with WB was the most advantageous test. PCR with plasma, which was developed locally, was the lengthiest (2 days) and most cumbersome procedure, although it had the lowest reagent costs. The hands-on time of HC with WB was similar to that of PCR with PBLs, but the procedure for the latter test was more complex and required more skills. The durations of HC with WB and PCR with PBLs were 7 and 10 h, respectively. Hence, HC with WB was a same-day test, whereas PBL PCR was a 2-day assay, which might affect clinical decision making. Overall, the costs per test were similar for PCR with plasma and HC with WB but higher for PCR with PBLs. The reimbursements obtained by using current procedural terminology codes would be the same for the three assays. Detection of CMV DNA for the purpose of triggering preemptive therapy is gradually supplanting CMV culture. This limits the availability of viral isolates for drug resistance testing, which is necessary for the clinical care of selected patients (14). This potential disadvantage may be overcome by continuing efforts to develop comprehensive genotyping assays for detection of CMV drug resistance (4, 5, 7, 15, 23). The low incidence of CMV disease in BMT recipients reported here and in other studies (18, 20) attests to the efficacy of CMV surveillance based on molecular techniques. Furthermore, in this study, detection of CMV DNA in WB might have completely prevented CMV disease in our patients, since in both patients who developed CMV disease their CMV reactivations were missed by PCR with plasma but were detected by HC with WB at least 1 week prior to the onset of symptoms. The data on comparative sensitivities taken together with the clinical observations suggest that further improvements in prevention of CMV disease in BMT patients might be possible by using WB or PBLs instead of plasma as the primary surveillance specimen. ACKNOWLEDGMENTS We thank Shaw Yi Kao and Irma Alfaro for kindly providing us the Roche and Digene kits, respectively; Lisa Michel for technical assistance; and Myron Levin for critical review of the manuscript.

4206

WEINBERG ET AL. REFERENCES

1. Boeckh, M., T. A. Gooley, and R. A. Bowden. 1998. Effect of high-dose acyclovir on survival in allogeneic marrow transplant recipients who received ganciclovir at engraftment or for cytomegalovirus pp65 antigenemia. J. Infect. Dis. 178:1153–1157. 2. Caliendo, A. M., K. St. George, J. Allega, A. C. Bullotta, L. Gilbane, and C. R. Rinaldo. 2002. Distinguishing cytomegalovirus (CMV) infection and disease with CMV nucleic acid assays. J. Clin. Microbiol. 40:1581–1586. 3. Cathomas, G., P. Morris, K. Pekle, I. Cunningham, and D. Emanuel. 1993. Rapid diagnosis of cytomegalovirus pneumonia in marrow transplant recipients by bronchoalveolar lavage using the polmerase chain reaction, virus culture, and the direct immunostaining of alveolar cells. Blood 81:1909– 1914. 4. Chou, S., N. S. Lurain, A. Weinberg, G. Y. Cai, P. L. Sharma, C. S. Crumpacker, and the Adult AIDS Clinical Trials Group CMV Laboratories. 1999. Interstrain variation in the human cytomegalovirus DNA polymerase sequence and its effect on genotypic diagnosis of antiretroviral drug resistance. Antimicrob. Agents Chemother. 43:1500–1502. 5. Chou, S., R. H. Waldemer, A. E. Senters, K. S. Michels, G. W. Kemble, R. C. Miner, and W. L. Drew. 2002. Cytomegalovirus UL97 phosphotransferase mutations that affect susceptibility to ganciclovir. J. Infect. Dis. 185:162–169. 6. Condie, R. M., and R. J. O’Reilly. 1984. Prevention of cytomegalovirus infection by prophylaxis with an intravenous, hyperimmune, native unmodified cytomegalovirus globulin. Randomized trial in bone marrow transplant recipients. Am. J. Med. 76:134–141. 7. Erice, A., C. Gil-Roda, J. L. Perez, H. H. Balfour, Jr., K. J. Sannerud, M. N. Hanson, G. Boivin, and S. Chou. 1997. Antiviral susceptibilities and analysis of UL97 and DNA polymerase sequences of clinical cytomegalovirus isolates from immunocompromised patients. J. Infect. Dis. 175:1087–1092. 8. Fox, J. C., P. D. Griffiths, and V. C. Emery. 1992. Quantification of human cytomegalovirus DNA using the polymerase chain reaction. J. Gen. Virol. 73:2405–2408. 9. Gerna, G., M. Furione, F. Baldanti, and A. Sarasini. 1994. Comparative quantitation of human cytomegalovirus DNA in blood leukocytes and plasma of transplant and AIDS patients. J. Clin. Microbiol. 32:2709–2717. 10. Goodrich, J. M., M. Boeckh, and R. Bowden. 1994. Strategies for the prevention of cytomegalovirus disease after marrow transplantation. Clin. Infect. Dis. 19:287–298. 11. Goodrich, J. M., M. Mori, C. A. Gleaves, C. Du Mond, M. Cays, D. F. Ebeling, W. C. Buhles, B. DeArmond, and J. Meyers. 1991. Early treatment with ganciclovir to prevent cytomegalovirus disease after allogeneic bone marrow transplantation. N. Engl. J. Med. 325:1601–1607. 12. Goodrich, J. M., R. A. Bowden, L. Fisher, C. Keller, G. Schoch, and J. D. Meyers. 1993. Ganciclovir prophylaxis to prevent cytomegalovirus disease after allogeneic marrow transplant. Ann. Intern. Med. 118:173–178. 13. Hebart, H., A. Schroder, J. Loffler, T. Klingebiel, H. Martin, B. Wassmann, F. Gerneth, H. Rabenau, G. Jahn, L. Kanz, C. A. Muller, and H. Einsele. 1997. Cytomegalovirus monitoring by polymerase chain reaction of whole blood samples from patients undergoing autologous bone marrow or peripheral blood progenitor cell transplantation. J. Infect. Dis. 175:1490–1493. 14. Limaye, A. P., L. Corey, D. M. Koelle, C. L. Davis, and M. Boeckh. 2000. Emergence of ganciclovir-resistant cytomegalovirus disease among recipients of solid-organ transplants. Lancet 356:645–649. 15. Lurain, N. S., A. Weinberg, C. S. Crumpacker, and S. Chow. 2001. Sequencing of cytomegalovirus UL97 gene for genotypic antiviral resistance testing. Antimicrob. Agents Chemother. 45:2775–2780.

J. CLIN. MICROBIOL. 16. Meyers, J. D. 1986. Infection in bone marrow transplant recipients. Am. J. Med. 81:27–38. 17. Meyers, K. D., P. Ljungman, and L. D. Fisher. 1990. Cytomegalovirus excretion as predictor of cytomegalovirus disease after marrow transplantation: importance of cytomegalovirus viremia. J. Infect. Dis. 162:373–380. 18. Nguyen, Q., R. Champlin, S. Giralt, K. Rolston, I. Raad, K. Jacomson, C. Ippoliti, D. Hecht, J. Tarrand, M. Luna, and E. Whimbey. 1998. Late cytomegalovirus pneumonia in adult allogeneic blood and marrow transplant recipients. Clin. Infect. Dis. 28:618–623. 19. Razonable, R. R., R. Brown, J. Wilson, C. Groettum, W. Kemers, M. Espy, F. Thomas, and C. Paya. 2002. The clinical use of various blood compartments for cytomegalovirus (CMV) DNA quantitation in transplant recipients with CMV disease. Transplantation 73:968–973. 20. Reusser, P., H. Einsele, J. Lee, L. Volin, M. Rovira, D. Engelhard, J. Finke, C. Cordonnier, H. Link, and P. Ljungman for the Infectious Diseases Working Party of the European Group for Blood and Marrow Transplantation. 2002. Randomized multicenter trial of foscarnet versus ganciclovir for preemptive therapy of cytomegalovirus infection after allogeneic stem cell transplantation. Blood 99:1159–1164. 21. Roberts, T. C., D. C. Brennan, R. S. Buller, M. Gaudreault-Keener, M. A. Schnitzler, K. E. Sternahell, K. A. Garlock, G. G. Singer, and G. A. Storch. 1998. Quantitative polymerase chain reaction to predict occurrence of symptomatic cytomegalovirus infection and assess response to ganciclovir therapy in renal transplant recipients. J. Infect. Dis. 178:626–635. 22. Schmidt, G. M., D. A. Horak, J. C. Nilan, S. R. Duncan, S. J. Forman, J. A. Zaia, and the City of Hope-Stanford-Syntex CMV Study Group. 1991. A randomized, controlled trial of prophylactic ganciclovir for cytomegalovirus pulmonary infection in recipients of allogeneic bone marrow transplants. N. Engl. J. Med. 324:1005–1011. 23. Spector, S. A., K. Hsia, D. Wolf, M. Shinkai, and I. Smith. 1995. Molecular detection of human cytomegalovirus and determination of genotypic ganciclovir resistance in clinical specimens. Clin. Infect. Dis. Suppl. 2:S170–S173. 24. Verdonck, L. F., A. W. Dekker, J. Rozenberg-Arska, and M. R. van den Hoek. 1997. A risk-adapted approach with a short course of ganciclovir to prevent cytomegalovirus (CMV) pneumonia in CMV-seropositive recipients of allogeneic bone marrow transplants. Clin. Infect. Dis. 24:901–907. 25. Weinberg, A., D. Spiers, G. Y. Cai, C. M. Long, and R. Sun. 1998. Evaluation of commercial PCR kit for diagnosis of cytomegalovirus infection of central nervous system. J. Clin. Microbiol. 36:3382–3384. 26. Weinberg, A., and S. Li. 1999. Evaluation of PCR primers for cytomegalovirus detection. J. Clin. Microbiol. 37:1658–1659. 27. Weinberg, A., T. N. Hodges, S. Li, G. Y. Cai, and M. R. Zamora. 2000. Comparison of PCR, antigenemia assay, and rapid blood culture for detection and prevention of cytomegalovirus disease after lung transplantation. J. Clin. Microbiol. 38:768–772. 28. Wingard, J. R., S. Piantadosi, W. H. Burns M. L. Zahurak, G. W. Santos, and R. Saral. 1990. Cytomegalovirus infections in bone marrow transplant recipients given intensive cytoreductive therapy. Rev. Infect. Dis. 12:S793– S804. 29. Zaia, J. A., G. M. Gallez-Hawkins, B. R. Tegetmeier, A. Ter Veer, X. Li, J. C. Niland, and S. J. Forman. 1997. Late cytomegalovirus disease in marrow transplantation is predicted by virus load in plasma. J. Infect. Dis. 176:782– 785. 30. Zipeto, D., S. Morris, C. Hong, A. Dowling, R. Wolitz, T. C. Merigan, and L. Rasmussen. 1995. Human cytomegalovirus (CMV) DNA in plasma reflects quantity of CMV DNA present in leukocytes. J. Clin. Microbiol. 33:2607– 2611.