Viral Infections Cytomegalovirus viral load monitoring after ... - Nature

9 downloads 0 Views 125KB Size Report
commonly after allogeneic BMT, even in patients receiv- ing antiviral prophylaxis. .... (1) ever positive, (2) maximum CMV viral load, (3) the occurrence of a CMV ...
Bone Marrow Transplantation (2003) 32, 795–800 & 2003 Nature Publishing Group All rights reserved 0268-3369/03 $25.00

www.nature.com/bmt

Viral Infections Cytomegalovirus viral load monitoring after allogeneic bone marrow transplantation in patients receiving antiviral prophylaxis BP Howden1, A Michaelides2, DW Spelman1,2, A Spencer3, AP Schwarer3, S Wesselingh1 and TC Kotsimbos4 1 Department of Infectious Diseases, Alfred Hospital, Commercial Rd., Prahran, Victoria, 3181, Melbourne, Australia; 2Department of Microbiology, Alfred Hospital, Commercial Rd., Prahran, Victoria, 3181, Melbourne, Australia; 3Bone Marrow Transplant Programme, Alfred Hospital, Commercial Rd., Prahran, Victoria, 3181, Melbourne, Australia; and 4Department of Respiratory Medicine, Alfred Hospital, Commercial Rd., Prahran, Victoria, 3181, Melbourne, Australia

Summary: Cytomegalovirus viral load measurement is a powerful new tool for monitoring of CMV disease; however, the optimal strategy for use is unknown. Weekly plasma CMV viral loads and CMV-related outcomes were monitored in 46 consecutive allogeneic bone marrow transplantation (BMT) recipients receiving standardised antiviral prophylaxis. A total of 412 CMV viral loads were quantitated in the first 100 days post transplantation with 77 positive samples (19%) in 20 patients (43%). No patient with all negative CMV viral load results developed CMV disease. Two of three patients with highly positive CMV viral loads (first positive p30 days post transplant, maximum viral load X5000 copies/ml, and X50% of samples positive) developed CMV disease. A total of 17 patients with positive CMV viral loads, who did not meet the criteria for highly positive, did not develop CMV disease. CMV viral load detection was higher in recipients who were CMV sero-positive. In conclusion, CMV disease did not occur in the setting of a persistently negative CMV viral load. A positive CMV viral load result occurred commonly after allogeneic BMT, even in patients receiving antiviral prophylaxis. Bone Marrow Transplantation (2003) 32, 795–800. doi:10.1038/sj.bmt.1704230 Keywords: cytomegalovirus; viral load

Cytomegalovirus (CMV) infection and disease remain one of the most important complications in recipients of allogeneic bone marrow and peripheral blood stem cell (PBSC) transplants.1 Interstitial pneumonitis is the most serious manifestation of CMV disease in these patients, with an associated mortality of 30–48% despite adequate

Correspondence: Dr BP Howden, Department of Microbiology, Austin and Repatriation Medical Centre, Locked Bag 25, Heidelberg 3084, Victoria, Australia. E-mail: [email protected] Received 03 March 2003; accepted 16 April 2003

treatment.2 Successful prevention and early detection of disease are particularly important in preventing morbidity and mortality from CMV in bone marrow transplantation (BMT) recipients. A number of groups at high risk for development of CMV disease have been identified. These include recipients of a graft from an unrelated donor, or a donor mismatched for one or more HLA class I or II alleles; patients with severe graft-versus-host disease (GVHD) who require additional immunosuppression leading to later onset CMV disease; a sero-positive recipient or receipt of a graft from a donor who is CMV seropositive.2,3 Routine diagnostic strategies for CMV disease have included viral culture and shell vial assays. However, these are inefficient and have poor sensitivity. The CMV antigenaemia assay is currently widely used to monitor BMT recipients,4 but there are some problems associated with this method. The increased specificity of these assays for CMV disease episodes is associated with decreased sensitivity for CMV detection. Secondly, it may be difficult to perform the assay before engraftment because of low leucocyte numbers.4 The development of CMV PCR assays has improved efficiency and sensitivity for direct detection of CMV in various specimens, including blood, plasma, urine, CSF and biopsy material. Recent reports of quantitative CMV PCR measuring CMV DNA viral load suggest that this is a powerful new tool which improves the specificity of CMV disease diagnosis, and also improves surveillance of patients post transplantation to predict likelihood of subsequent CMV disease, and hence allow appropriate use of pre-emptive therapy.5–7 Current strategies for the prevention of CMV disease include ganciclovir prophylaxis, or ganciclovir pre-emptive therapy based on detection of antigenaemia or CMV DNA.8 The optimal strategy for CMV viral load monitoring after BMT, and interpretation of results in patients receiving antiviral prophylaxis is unknown. The aims of this study were to (1) describe the CMV viral load in the plasma of a cohort of BMT recipients receiving antiviral prophylaxis during the first 100 days post transplantation; (2) assess the impact of antiviral prophylaxis on CMV viral load; (3) assess the association between

CMV viral load monitoring after BMT

BP Howden et al

796

CMV viral load dynamics and CMV-related outcomes, including CMV disease, acute and chronic GVHD and death.

intravenous ganciclovir (5 mg/kg twice daily) with or without CMV immunoglobulin. Foscarnet was available as an alternative agent.

CMV viral loads Patients and methods Patients and immunosuppression A prospective cohort study of all patients who had an ablative or nonmyeloablative BMT or PBSC transplant was performed at The Alfred Hospital between 1 March 2000 and the 31 May 2001. All patients had routine monitoring of CMV viral load post transplantation and were included in this study. Weekly CMV viral loads were performed until day 100 after transplantation where possible, and monitoring for CMV disease was also undertaken for 100 days post transplant. Patients were also followed until January 2003 to monitor chronic GVHD and death. Data were collected regarding demographic factors, reason for transplantation, donor and recipient CMV status, immunosuppressive therapy, HLA matching, antiviral prophylaxis, GVHD and occurrence of CMV disease. Definitions of GVHD were based on recognised criteria.9 All patients received cyclosporin post transplantation. Some patients received cyclosporin alone as GVHD prophylaxis, after conditioning with total body irradiation and high-dose cyclophosphamide, if they had a matched sibling donor and were at high risk of relapse. Patients who had a miniallograft were given fludarabine as conditioning. Equine antithymocyte globulin was used for all unrelated allografts before and after stem cell infusion, and in mismatched related grafts as conditioning. Since January 2001, equine antithymocyte globulin has also been used as conditioning for all miniallografts.

Antiviral prophylaxis and treatment Prophylaxis against CMV disease was used as follows. A CMV sero-positive recipient was given intravenous ganciclovir prior to transplantation. After transplantation, oral valacyclovir was used as herpes simplex virus prophylaxis (500 mg daily) until engraftment, when it was ceased, at which time intravenous ganciclovir (5 mg/kg 3 times per week) was recommenced and continued until day 84 post transplant. Ganciclovir was started if the neutrophil count was above 1.0  109/l and platelet count above 20  109/l on day 14 post PBSC transplant, or day 21 post BMT. CMV sero-negative recipients were not given ganciclovir prior to transplantation, but commenced valacyclovir after transplant. If the donor was CMV sero-positive, intravenous ganciclovir was commenced at engraftment and continued until day 84 post transplant, as above. If the donor was also CMV sero-negative, then ganciclovir was not used. Patients intolerant of ganciclovir (usually low blood counts) received high-dose valacyclovir (2 g four times daily) as CMV prophylaxis. Full-dose pre-emptive therapy with ganciclovir was usually used if the CMV viral load was above 10 000 copies/ml or if the patient was at high risk for CMV disease (unrelated donor, GVHD and significant immunosuppression). CMV disease was treated with Bone Marrow Transplantation

CMV viral loads were assessed in plasma samples using the commercially available semiautomated COBAS Amplicor CMV monitor test (Roche diagnostics systems, Branchburg NJ, USA). Testing of samples was performed according to the manufacturer’s instructions. Plasma samples were recorded as copies/ml. The lower detection limit of the assay is 400 copies/ml.

CMV viral load measurements CMV viral load results for each patient were recorded as (1) ever positive, (2) maximum CMV viral load, (3) the occurrence of a CMV viral load X5000 copies/ml and (4) the proportion of positive results per patient. Maximum CMV viral load and a CMV viral load 45000 copies/ml have previously been shown to be associated with an increased risk of CMV disease.5,6 We chose to define highly positive CMV viral load results as all three of the following occurring together; first positive p30 days post transplant, with maximum CMV viral load X5000 copies/ml, and X50% of samples positive. Patients with a positive CMV viral load result who did not fit the criteria for highly positive were considered to have intermediate CMV results.

CMV disease The definition of CMV disease was based on criteria defined by the International Workshop on CMV Disease.10 The case definition for gastrointestinal CMV disease was gastrointestinal symptoms plus histological evidence of CMV from gut biopsies. Diagnostic procedures such as bronchoscopy and endoscopy were performed when clinically indicated.

Statistical analysis Statistical analysis was performed using Stata version 5.0 (Statacorp 1997; Stata Statistical Software, Release 5.0, Stata corporation, College Station, TX, USA). Univariate analysis was performed using the w2 test and Fisher’s exact test.

Results Patient characteristics A total of 46 patients received an allogeneic bone marrow or PBSC transplant between March 2000 and May 2001 (30 myeloablative, 16 nonmyeloablative). Information was available on all patients and details are presented in Table 1. Acute lymphocytic leukaemia was the most common indication for transplantation. All patients received valacyclovir and 40 patients (87%) received prophylactic ganciclovir. Six patients did not receive prophylactic ganciclovir because the donor and the

CMV viral load monitoring after BMT BP Howden et al

797 Patient characteristics

Age Mean Range Transplant Type Related Unrelated Conditioning Myeloablative Nonmyeloablative Matching Matched Unmatched Diagnosis Acute leukaemia Non-Hodgkin’s lymphoma Chronic leukaemia Other CMV status D+R+ D+R DR+ DR Ganciclovir prophylaxis Yes No

46 30 16

40.7 17–61

Log CMV Viral Load (copies/ml)

Total number of patients Sex Male Female

CMV Viral Load (Patient 1) 6 GCV. CMVIG

F.N.V.D

5 4 3 2 1 0 1

2

3

29 17

15 6 6 19 20 8 12 6

5

6

7

8

9

10

CMV Viral Load (Patient 2)

30 16 40 6

4

Weeks post transplant

Log CMV Viral Load (copies/ml)

Table 1

6

V.AP.Dys

F.D

5

GCV

4 3 2 1 0 1

2

3

4

5

6

7

8

9 10 11 12 13 14

Weeks post transplant Figure 1 CMV viral load results, symptoms and therapy in patients with

40 6

recipient were CMV sero-negative. Two patients had symptomatic, biopsy-proven CMV disease in the first 100 days post transplant. There were 15 deaths (33%) in the first 100 days of follow-up, and 12 (26%) after 100 days post transplant. Acute GVHD developed in 22 patients (48%) and chronic GVHD developed in 20 patients (43%).

CMV detection In total, 412 CMV viral load tests were performed in the 46 patients with 77 positive results (19%) in 20 patients. A total of 26 patients had all negative CMV viral load results throughout the period of follow-up and no CMV disease was evident in this patient group.

CMV viral load In all, 20 patients (43%) had a positive result for CMV at some time during the 100 days at follow-up. Three patients had highly positive CMV viral load results, with two of these patients developing biopsy-proven gastrointestinal CMV disease. In the patients with a positive CMV viral load, the mean maximum viral load was 21 510 copies/ml (range 400–100 000 copies/ml). A total of 11 patients had a maximum CMV viral load X5000 copies/ml, but only two patients developed CMV disease (Table 2). In those who had a positive CMV viral load result and did not develop CMV disease, the first positive occurred at a median of 48.5 days post transplant (range 21–89 days). The two patients who developed CMV disease had a positive CMV viral load earlier, at 19 and 25 days post transplant.

CMV disease. F ¼ fever; N ¼ nausea; V ¼ vomiting; D ¼ diarrhoea; GCV ¼ intravenous ganciclovir; CMVIG ¼ CMV immunoglobulin; AP ¼ abdominal pain; Dys ¼ dysphagia.

Patients with CMV disease Two patients (4.3%) developed histopathologically proven CMV disease during 100 days of follow-up after transplantation. The details of these patients are presented in Figure 1. Both patients had prolonged high levels of CMV in the plasma, which persisted despite treatment with ganciclovir. Patient 1 developed gastrointestinal symptoms 2 weeks after transplantation. A biopsy of sigmoid colon 3 weeks post transplant was normal, and a gastric biopsy 5 weeks post transplant demonstrated nonspecific inflammation. Both the donor and recipient were CMV sero-positive and the patient received intravenous ganciclovir prophylaxis prior to transplantation. The CMV viral load became positive at the time of symptom onset. However, the diagnosis was not confirmed by biopsy until six weeks post transplant on gastric and sigmoid colon specimens. The CMV viral load remained high despite treatment with intravenous ganciclovir and CMV immune globulin, and the patient died 8 weeks post transplant. The pattern for patient 2 was similar. The donor was CMV sero-negative, the recipient CMV sero-positive and intravenous ganciclovir was given prior to transplantation. Symptoms of gastrointestinal disease started 3–4 weeks post transplant, at which time the CMV viral load became positive. Sigmoid biopsies at weeks four and six post transplant did not demonstrate CMV disease. A gastric biopsy confirmed the diagnosis 10 weeks post transplant and treatment with intravenous ganciclovir Bone Marrow Transplantation

CMV viral load monitoring after BMT

BP Howden et al

798 Table 2

Summary of results, symptoms and treatment for all patients who had a positive CMV viral load result after BMT

Patient

GCV prophylaxis

Viral load

Symptoms

Days to first positive

Number of samples

Number positive (%)

Yes

19

8

88

49 200

F,N,V,D

Yes

25

12

75

19 100

F,D,V,AP

Yes Yes

26 31

4 3

25 33

48 500 1800

5 6 7 8 9

Yes Yes No Yes Yes

21 83 64 81 40

13 15 8 12 15

85 20 13 8 27

68 200 10 600 400 400 6510

10

Yes

55

10

30

30 700

11 12 13 14

Yes Yes Yes Yes

60 65 24 46

15 15 18 11

7 40 44 45

915 30 700 4100 000 4730

15

Yes

49

9

33

583

16 17 18 19

Yes Yes Yes Yes

48 46 66 38

12 9 5 9

50 33 20 11

28 500 26 100 400 627

20

Yes

89

13

15

2230

CMV disease 1 2 No CMV disease 3 4

Treatment

VL response

Peak (copies/ml) IV GCV, CMVIG IV GCV

None

F,C, F, pulm infiltrate — — — — LOW, N, V Raised LFT’s — — — —

IV GCV None

No further test No further test

Oral GCV IV GCV — — —

Negative Negative Negative Negative Negative

IV GCV

Negative 2 weeks

— IV GCV IV GCV —

GI ulceration — — — F, TCP

IV GCV, CMVIG — — — Foscarnet, CMVIG —

Negative 1 week Negative 3 weeks Negative 10 days Positive 3 weeks, no further tests Negative 1 week



None

3 weeks 10 days 1 week 1 week 2 weeks

Negative 3 weeks No further test Negative 1 week Negative 1 week No further test

NOTE: F ¼ fever; N ¼ nausea; V ¼ vomiting; D ¼ diarrhoea; IV GCV ¼ intravenous ganciclovir; CMVIG ¼ CMV immunoglobulin; AP ¼ abdominal pain; C ¼ cough; TCP ¼ thrombocytopaenia.

was commenced. The CMV viral load remained high despite the ganciclovir, and although symptoms improved the patient died 2 months later.

CMV-positive patients without proven CMV disease A total of 18 patients had a positive viral load result without definite evidence of CMV disease. Table 2 summarises the viral load results for all patients with a positive CMV viral load. The majority of these patients had a small number of positive results, or had a fluctuating pattern of positive and negative results. Persistent positive viral load results were not a feature of this group, although a number of these patients had high viral load results without evidence of disease (4100 000 copies/ml in one patient). For a number of patients who received empiric treatment with intravenous ganciclovir or foscarnet, the CMV viral load responded quickly in most cases, with a negative result detected within 2–3 weeks for all patients from the time of commencing treatment.

Univariate analysis A univariate analysis assessing factors associated with a positive CMV viral load result and a CMV viral load X5000 copies/ml was performed, including the variables Bone Marrow Transplantation

transplant type, HLA matching, donor relatedness, acute GVHD and CMV donor and recipient sero-status. It showed that transplant type, HLA matching, donor relatedness and acute GVHD were not significantly associated with CMV viral load results. Donor and recipient CMV sero-status was associated with CMV viral load results. A CMV sero-positive donor and recipient was significantly associated with a CMV viral load result X5000 copies/ml (P ¼ 0.03), whereas a transplant from a sero-negative donor to a sero-positive recipient was associated with any CMV viral load detection and approached statistical significance (P ¼ 0.07). A univariate analysis of CMV viral load results (any detection, level X5000 copies/ml, highly positive results) with the outcomes of death in the first 100 days post transplant, death greater than 100 days post transplant, acute GVHD and chronic GVHD was performed. There were insufficient numbers of proven CMV disease cases to include this outcome in the statistical analysis. CMV viral load results did not predict for death in the first 100 days post transplant, acute or chronic GVHD, or all CMVrelated outcomes when pooled. A viral load result X5000 copies/ml was associated with death more than 100 days post transplant (P ¼ 0.01); however, seven out of 12 of these deaths were related to relapse and progression of disease.

CMV viral load monitoring after BMT BP Howden et al

799

Discussion CMV disease remains an important and life-threatening complication of bone marrow and PBSC transplantation. CMV viral load monitoring offers the best opportunity for risk stratification in these patients, and allows the use of pre-emptive therapy and monitoring of the response to therapy. The use of ganciclovir prophylaxis after bone marrow transplantation may alter the patterns of CMV viral load results and indications for pre-emptive therapy may be different. The CMV viral load dynamics in a cohort of patients receiving standard antiviral prophylaxis after bone marrow transplantation have not been previously reported. This study will help to guide the interpretation of CMV viral load results in patients after BMT. The 46 patients in this study are a consecutive, representative sample of our patient group over a 15month period. CMV disease occurred in two patients (4.3%) in the first 100 days post transplant, which is a lower rate compared to previous reports.5,11,12 A large number of CMV viral load tests were performed in the first 100 days post transplant. Interestingly, almost half the patients had a positive CMV viral load result during the study, despite the regular use of antiviral prophylaxis. Gor et al reported a similar rate of 45% of patients with a positive viral load result after BMT in patients receiving acyclovir prophylaxis.5 Importantly, most patients who had a positive viral load did not develop clinically overt CMV disease. The two patients with proven CMV disease had persistent, high positive viral load results, which has been a feature in previous reports of CMV viral load testing in transplant recipients.6,13 A persistently negative CMV viral load result has excellent negative predictive value for CMV disease in patients who receive antiviral prophylaxis after BMT. A single positive CMV viral load result alone has low predictive value for CMV disease. In the two patients with CMV gastrointestinal disease, the CMV viral load became positive around the time of symptom onset and remained high until the diagnosis was confirmed. Of interest, the CMV viral load remained persistently high in these two patients despite fulldose intravenous ganciclovir, and one of these patients died soon after. CMV viral load levels have previously been shown to fall after commencement of intravenous ganciclovir in lung transplant recipients, although the time to response is not clear.13 Patients in this study without CMV disease who received treatment doses of ganciclovir or foscarnet had an undetectable CMV viral load within 2–3 weeks. The persistence of a high viral load result despite therapy may be a reason to consider alternate or additional therapy in this patient group, particularly if prophylactic ganciclovir has been used previously in the patient. Ganciclovir resistance testing was not performed in these two patients, although development of resistance was a possibility. Resistance testing is now recommended in our BMT patients who have persistent positive CMV viral loads despite ganciclovir therapy. Although foscarnet may have been appropriate in these two patients, it was not used. The main limitation of this study is the small number of patients who had proven CMV disease. This made

statistical analysis of CMV viral load and disease impossible. The need for histopathological evidence to confirm the diagnosis of CMV disease remains a limitation of this type of study. Some patients may have had symptoms attributable to CMV, but did not have biopsyproven evidence of disease. Nevertheless, our study results clearly indicate that there are three subgroups of patients regarding CMV reactivation post BMT. Those with persistent negative CMV viral load results in whom CMV disease does not appear to occur, those with highly positive CMV viral load results in whom CMV disease does occur and those with indeterminate CMV viral load results, where correlation with reactivation and disease is not clear. We have not demonstrated a significant difference in the CMV viral load in patients based on known risk factors for CMV disease, except with regard to donor and recipient CMV status. A positive recipient from a negative donor had a higher rate of CMV viral load detection that approached statistical significance. Recipient CMV seropositivitiy has previously been shown to be a risk factor for CMV pneumonitis.2,7 If the donor and recipient were seropositive, there was an increased rate of CMV viral load detection X5000 copies/ml. This finding is contrary to previous studies that have shown that sero-positive recipients of a bone marrow transplant had a lower viral load and lower risk of CMV disease if the donor was also sero-positive.5,14 The long-term follow-up in this study allowed correlation of death after 100 days post transplant, and chronic GVHD with CMV viral load results in the first 100 days post transplant. The optimal approach to CMV disease prevention after BMT is debated, and the effect of some CMV reactivation without overt disease is unclear. In this study, any CMV detection post transplant was not associated with death or chronic GVHD, although patients with a CMV viral load result X5000 copies/ml had a higher rate of death after 100 days. It is difficult to draw conclusions about the cause and effect of this association, particularly given the high rate of relapse and progressive disease in these patients; however, it warrants further study. The optimal frequency of CMV viral load monitoring after BMT is unknown, but we feel weekly monitoring is adequate. The two patients with proven CMV disease had positive CMV viral loads for a number of weeks before the diagnosis was obtained on biopsy. More frequent monitoring may be useful, particularly in high-risk patients, although this requires further study. In summary, we have described the changes in CMV viral load in the first 100 days after allogeneic bone marrow and PBSC transplantation in 46 consecutive patients who received antiviral prophylaxis. The findings in this study will help to guide the use of this powerful diagnostic tool in patients after BMT. Persistently negative CMV viral load has excellent negative predictive value for CMV disease. A single positive result is not clinically useful, although a persistent high viral load was associated with severe CMV disease in two patients. Further studies are required of patients with intermediate CMV viral load results to correlate these results with short- and long-term clinical outcomes. Bone Marrow Transplantation

CMV viral load monitoring after BMT

BP Howden et al

800

References 1 Einsele H, Hebart H. Cytomegalovirus infection following stem cell transplantation. Haematologica 1999; 84 (Suppl. EHA-4): 46–49. 2 Enright H, Haake R, Weisdorf D et al. Cytomegalovirus pneumonia after bone marrow transplantation. Risk factors and response to therapy. Transplantation 1993; 55: 1339–1346. 3 Miller W, Flynn P, McCullough J et al. Cytomegalovirus infection after bone marrow transplantation: an association with acute graft-v-host disease. Blood 1986; 67: 1162–1167. 4 Machida U, Kami M, Fukui T et al. Real-time automated PCR for early diagnosis and monitoring of cytomegalovirus infection after bone marrow transplantation. J Clin Microbiol 2000; 38: 2536–2542. 5 Gor D, Sabin C, Prentice HG et al. Longitudinal fluctuations in cytomegalovirus load in bone marrow transplant patients: relationship between peak virus load, donor/recipient serostatus, acute GVHD and CMV disease. Bone Marrow Transplant 1998; 21: 597–605. 6 Emery VC, Sabin CA, Cope AV et al. Application of viralload kinetics to identify patients who develop cytomegalovirus disease after transplantation. Lancet 2000; 355: 2032–2036. 7 Qamruddin AO, Oppenheim BA, Guiver M et al. Screening for cytomegalovirus (CMV) infection in allogeneic bone marrow transplantation using a quantitative whole blood polymerase

Bone Marrow Transplantation

8

9

10

11

12

13

14

chain reaction (PCR) method: analysis of potential risk factors for CMV infection. Bone Marrow Transplant 2001; 27: 301–306. Boeckh M. Current antiviral strategies for controlling cytomegalovirus in hematopoietic stem cell transplant recipients: prevention and therapy. Transplant Infect Dis 1999; 1: 165–178. Przepiorka D, Weisdorf D, Martin P et al. 1994 Consensus Conference on Acute GVHD Grading. Bone Marrow Transplant 1995; 15: 825–828. Ljungman P, Plotkin SA. Workshop of CMV disease: definitions, clinical severity scores and syndromes. Scand J Infect Dis-Suppl 1995; 99: 87–89. Humar A, O’Rourke K, Lipton J et al. The clinical utility of CMV surveillance cultures and antigenemia following bone marrow transplantation. Bone Marrow Transplant 1999; 23: 45–51. Ljungman P, Aschan J, Lewensohn-Fuchs I et al. Results of different strategies for reducing cytomegalovirus-associated mortality in allogeneic stem cell transplant recipients. Transplantation 1998; 66: 1330–1334. Michaelides A, Liolios L, Glare EM et al. Increased human cytomegalovirus (HCMV) DNA load in peripheral blood leukocytes after lung transplantation correlates with HCMV pneumonitis. Transplantation 2001; 72: 141–147. Grob JP, Grundy JE, Prentice HG et al. Immune donors can protect marrow transplant recipients from severe cytomegalovirus infections. Lancet 1987; 1: 774–776.