Targeted Individual Prophylaxis Offers Superior Risk Stratification for Cytomegalovirus Reactivation After Liver Transplantation

Transcription

1 LIVER TRANSPLANTATION 21: , 2015 ORIGINAL ARTICLE Targeted Individual Prophylaxis Offers Superior Risk Stratification for Cytomegalovirus Reactivation After Liver Transplantation Siddharth Sood, 1,2,3 Craig Haifer, 1 Lijia Yu, 3 Julie Pavlovic, 1 Paul J. Gow, 1 Robert M. Jones, 1 Kumar Visvanathan, 3 Peter W. Angus, 1 and Adam G. Testro 1 1 Liver Transplant Unit Victoria, Austin Health, 2 Department of Gastroenterology and Hepatology, Royal Melbourne Hospital, and 3 Innate Immune Laboratory, St. Vincent s Hospital, University of Melbourne, Melbourne, Australia Cytomegalovirus (CMV) can reactivate following liver transplantation. Management of patients currently considered low risk based on pretransplant serology remains contentious, with universal prophylaxis and preemptive strategies suffering from significant deficiencies. We hypothesized that a CMV-specific T cell assay performed early after transplant as part of a preemptive strategy could better stratify low-risk (recipient seropositive) patients. We conducted a prospective, blinded, observational study in 75 adult recipients. QuantiFERON-cytomegalovirus was performed both before and at multiple times after transplant. Low-risk patients (n 5 58) were monitored as per unit protocol and treatment was commenced if CMV > 1000 copies/ml (DNAemia). Twenty patients needed antiviral treatment for other reasons and were censored (mainly for rejection or herpes simplex virus infection); 19/38 (50%) of the remaining low-risk patients developed DNAemia at mean 34.6 days after transplant. A week 2 result of <0.1 IU/mL was significantly associated with risk of subsequent DNAemia (hazard ratio [HR], 6.9; P ). The positive predictive value of 80% suggests these patients are inappropriately labeled low risk and are actually at high likelihood of CMV reactivation. A secondary cutoff of <0.2 IU/mL was associated with moderate risk (HR, 2.8; P ). In conclusion, a protocol based on a single early CMV-specific T cell based assay would offer improved risk stratification and individualization of patient management after transplant. This could offer improved drug and service utilization and potentially result in significant improvements over both currently used protocols to manage supposedly low-risk patients. Liver Transpl 21: , VC 2015 AASLD. Received April 14, 2015; accepted June 20, See Editorial on Page 1462 Cytomegalovirus (CMV) has a benign primary infection followed by a prolonged dormant phase where the virus is controlled by cell-mediated immunity with strong virus-specific cytotoxic T cell responses in immune-competent individuals. 1-3 Most adults (50%- 85%) have been previously infected with CMV. 4 Current immunosuppression used to prevent organ rejection weakens immune responses, resulting in increased susceptibility to CMV reactivation 5 which is associated with increased risk of graft loss or death after liver transplantation (LT). 6 Apart from direct Abbreviations:, seronegative; 1, seropositive; ACR, acute cellular rejection; CI, confidence interval; CMV, cytomegalovirus; D, donor; DNAemia, CMV>1000 copies/ml; ELISPOT, enzyme-linked immunosorbent spot; HR, hazard ratio; HSV, herpes simplex virus; IFNg, interferon g; LT, liver transplantation; MIT, mitogen; PCR, polymerase chain reaction; PHA, phytohaemagglutin; PPV, positive predictive value; QFN-CMV, QuantiFERON-cytomegalovirus; QFN-MIT, QuantiFERON-mitogen; QFN, QuantiFERON; R, recipient. This work received sponsorship by Qiagen, USA, and was partly funded by a grant from the Austin Medical Research Foundation. Qiagen sponsored the study but was not involved in analysis or preparation of the manuscript. Lijia Yu was a prior employee of Cellestis, Ltd. (now owned by Qiagen, USA). Kumar Visvanathan and Lijia Yu are involved with other sponsored research studies by Qiagen, USA. Siddharth Sood, Adam G. Testro, Peter W. Angus, Paul J. Gow, Craig Haifer, Julie Pavlovic, Robert M. Jones: Nothing to report. Address reprint requests to Siddharth Sood, Department of Gastroenterology and Hepatology, Royal Melbourne Hospital, Level 3 Centre, Royal Parade, Parkville, Victoria, 3050, Australia. Telephone: ; FAX: ; siddharth.sood@mh.org.au DOI /lt View this article online at wileyonlinelibrary.com. LIVER TRANSPLANTATION.DOI /lt. Published on behalf of the American Association for the Study of Liver Diseases VC 2015 American Association for the Study of Liver Diseases.

2 LIVER TRANSPLANTATION, Vol. 21, No. 12, 2015 SOOD ET AL effects, CMV infection has been associated with an indirect increased risk of other infections (bacterial, fungal, and viral), acute cellular rejection (ACR), and vanishing bile duct syndrome. 7,8 Whether the increased graft failure and death are related to the viral infection or reflect an over suppressed immune system is unknown. Patients considered high risk for CMV reactivation based on donor (D) and recipient (R) pretransplant serology (D seropositive [1]/R seronegative [ ]) are commonly offered prophylaxis. Optimal management of patients stratified as low risk (D /R1, D1/R1) is more contentious. Despite being classified low risk, many patients still develop CMV when monitored as part of a preemptive strategy, which is also fraught with logistical monitoring issues and excessive pathology costs. Conversely, some units employ universal prophylaxis which is associated with high drug costs and increased risk of side effects, with many patients receiving unnecessary medication. Both strategies are considered viable approaches to managing posttransplant CMV reactivation according to international consensus guidelines, but some theorize that future approaches may include use of individualized immunodiagnostics to provide a superior outcome. 9 QuantiFERON-cytomegalovirus (QFN-CMV; Qiagen, Valencia, CA) is an immune assay that measures the T cell interferon g (IFNg) response to a variety of CMV proteins. 10 A nonreactive QFN-CMV is suggestive of an inadequate immune response toward the virus, and it is believed to highlight patients at risk of future CMV reactivation. The largest studies have so far focused on QFN-CMV in predicting late CMV (after prophylaxis has been given and ceased). 11 In this study, we aim to examine whether nonreactive QFN-CMV performed early after LT in low-risk patients is associated with risk of future CMV DNAemia. This improved ability to stratify individual patient risk based on more than just serology could allow targeted prophylaxis or reduced monitoring and could offer significant morbidity and cost benefits compared with current standards of care. PATIENTS AND METHODS Study Population Seventy-five adult LT patients were recruited as part of a prospective, blinded transplant immunity cohort study. QFN-CMV was performed before transplant and then regularlyaftertransplant(days,1,3,5;weeks,1,2;months, 1-3). Patients were followed for 6 months after transplant. Immunosuppression was per institution protocol, with a combination of corticosteroids (high-dose intravenous methylprednisolone for 5 days and then oral prednisolone on a slow wean during 3-6 months), a calcineurin inhibitor (either cyclosporine or tacrolimus), and either mycophenolate mofetil or azathioprine. Management of CMV Management of CMV was per unit protocol. High-risk patients (D1/R ) were given 6 months of prophylaxis with oral valganciclovir. Low-risk (D1/R1, D /R1) and very low-risk (D /R ) patients were monitored with CMV quantitative polymerase chain reaction (PCR) weekly or at each clinic appointment. CMV PCR was performed using a real-time PCR TaqMan (Thermo Fisher Scientific, Waltham, MA) assay. 12 As per unit protocol, if viral load increased above >1000 copies/ml, treatment with valganciclovir at 900 mg twice daily (renally adjusted) was commenced. After 2 successive months of undetectable viral load measurements, the dose was reduced to prophylaxis levels of 900 mg once daily until 6 months after transplant as per recommendations from the international consensus guidelines. 9 QFN-CMV Assay QFN-CMV was performed according to the manufacturer s instructions. The assay involved inserting 1 ml of whole blood into the QFN-CMV, QFN-nil (no antigen), and QFN-mitogen (MIT; phytohaemagglutin [PHA] MIT) Vacuette tubes (Greiner bio-one, Kremsm unster, Austria). Samples were incubated for 16 to 24 hours, following which plasma was harvested and enzymelinked immunosorbent assay was performed to measure IFNg levels as previously described. 10 The primary cutoff for a nonreactive result (suggestive of insufficient immunity and therefore risk of future CMV) was an absolute value of the QFN-CMV tube of <0.1 IU/mL, regardless of the results of the QFN-MIT or QFN-nil tubes. A secondary cutoff which is listed as the manufacturer s recommendation of <0.2 IU/mL was also investigated. Statistics Kaplan-Meier curves with the Mantel-Cox test were used to associate the QFN-CMV with future development of CMV DNAemia requiring commencement of prophylaxis according to unit protocol. Mean age is presented as 6 standard deviation, and patient groups were compared using the Mann-Whitney U test. Patients requiring antivirals for other reasons were censored, and as per international consensus guidelines, 9 those suffering ACR following reintroduction of pulsed corticosteroids began prophylaxis and were excluded from future analysis. A P value of <0.05 was considered significant. Data were analyzed using GraphPad Prism for Mac version 6.0 (GraphPad, La Jolla, CA). Institutional ethics were approved by Austin Health (HREC ), and this study complied with the national statement for ethical conduct in humans. All patients provided written informed consent, and no donor organs were obtained from executed prisoners or other institutionalized persons. RESULTS Seventy-five patients were recruited and prospectively followed after their LT (Fig. 1). The majority of patients were male (n 5 49, 65.3%) with a mean age of 50.0

3 1480 SOOD ET AL. LIVER TRANSPLANTATION, December 2015 Figure 1. Patient flow chart. (611.5) years old. Nine (12.0%) patients were very low risk (D /R ) with no prior exposure to CMV by either donor or recipient. All had QFN-CMV < 0.1 IU/mL before transplant and for the duration of the study period. One died from cardiac failure early after transplant, and none developed CMV DNAemia for the study duration. Eight (10.7%) patients were high risk for CMV and commenced prophylaxis (7 D1/R, 1 combined liverkidney transplant). Of the 7 patients who were D1/R, 4 (57.0%) still had a QFN-CMV < 0.1 IU/mL after 6 months of prophylaxis. A further 12 patients had clinical events before 2 weeks after transplant and began treatment: 2 patients had CMV detected at the time of transplant, 7 patients required reintroduction of pulsed steroids for rejection, and 3 patients required antiviral treatment for other infections (mainly herpes simplex virus [HSV]). The remaining 46 patients were monitored from 2 weeks after transplant for development of CMV DNAemia. Eight patients commenced antivirals for other reasons and were censored early: 3 suffered rejection needing pulsed steroids, 4 had other viral infections (HSV and shingles), and 1 commenced offprotocol antiviral treatment with a viral load of only 704 copies/ml (Fig. 1). This left 38 patients (50.7% of the cohort) monitored for risk of CMV reactivation. Despite positive pretransplant serology (R1), 21.0% had a QFN-CMV of <0.1 IU/mL before transplantation suggesting insufficient CD8 T cell response toward the virus. This level peaked at day 3, when the immune response appears to be at its weakest and when there is the highest proportion of patients having a nonreactive QFN-CMV (Fig. 2). Thereafter, there was a decrease in the percentage of patients with a nonreactive QFN-CMV until 2 weeks after transplant when the results began to plateau. Development of CMV DNAemia Nineteen of 38 (50.0%) monitored patients developed CMV DNAemia above the threshold of 1000 copies/ml at a median of 34 days after transplant (range, days) and commenced antiviral therapy. Median viral load at institution of therapy was 2527 copies/ml, with a median peak viral load of 3223 copies/ml. It took a median of 14 days for the viral load to become undetectable following the commencement of treatment. Only 2 patients developed CMV disease. There was no significant difference between the age of patients who developed CMV DNAemia ( years old) and those who were CMV-free ( years old; P ). Prior exposure of the donor to CMV made no impact to future CMV DNAemia, with 4 of 9 D /R1 patients requiring treatment. A recipient s pretransplant QFN-CMV (<0.1 IU/mL) was not associated with development of CMV DNAemia (5/8, 62.5%; hazard ratio [HR], 1.4; P ). Median IFNg production by the QFN-CMV assay appeared higher in patients who remained CMV-free at each time point throughout the study (Fig. 3).

4 LIVER TRANSPLANTATION, Vol. 21, No. 12, 2015 SOOD ET AL Figure 3. Median QFN in patients who suffered CMV DNAemia compared with those who were CMV-free (n 5 38). Figure 2. Percentage of R1 LT recipients with a nonreactive QFN-CMV (<0.1 IU/mL). The shaded area represents the group considered to have insufficient immune function toward CMV and at risk of reactivation. Ten of 38 (26.3%) R1 patients had cutoff levels below 0.1 IU/mL at week 2 after transplant, with 8 (80.0%) developing CMV DNAemia compared with 11 of 28 (39.3%) patients above the cutoff (Fig. 4; sensitivity 0.42; specificity 0.89). On Kaplan-Meier analysis, QFN- CMV 2 weeks after transplant was significantly associated with future risk of CMV DNAemia using the primary threshold of <0.1 IU/mL (Fig. 5A; HR, 6.9; P ). The higher cutoff of 0.2 IU/mL improved sensitivity but reduced specificity (sensitivity, 0.74; specificity, 0.53) and was also significantly associated with development of CMV DNAemia (Fig. 5B; HR, 2.8; P ), suggesting a moderate risk of CMV. QFN-CMV < 0.1 IU/mL measured at 1 week after transplant was also significantly associated with development of CMV DNAemia (HR, 3.1; P ), but a cutoff of 0.2 IU/mL was not (HR, 2.5; P ). The PHA MIT tubes were analyzed separately. Using similar cutoffs, a QFN-MIT < 0.1 IU/mL at 2 weeks after transplant was associated with a significant chance of developing CMV DNAemia (Fig. 5C; HR, 10; P ) with sensitivity of 0.32 and specificity of Manufacturer s guidelines suggest a QFN-MIT tube < 0.5 should be considered indeterminate. A majority of samples would have been excluded due to insufficient MIT response (11/19 patients who were CMV-free and 9/19 patients who had CMV DNAemia). The background reactivity (QFN-nil) tubes revealed no difference in patients who developed CMV (median, 0.06 IU/mL; 95% confidence interval [CI], ) compared with those who did not (median, 0.07 IU/mL; 95% CI, ). DISCUSSION The ideal management of CMV after LT remains debatable. On current guidelines, high risk is defined on pretransplant serology (D1/R ), and these patients are offered prophylaxis. Most adults have had prior exposure to CMV and are therefore considered low risk, but they still experience significant rates of CMV reactivation after transplant. Management of these recipients remains particularly controversial. Low-risk patients in this study (excluding those needing antiviral treatment for other reasons or suffering acute rejection) still had up to a 50% chance of developing CMV DNAemia. This is perhaps why many units employ a universal prophylaxis protocol. This strategy involves universal prophylaxis to all patients for 3 to 6 months, and it is common throughout North America, with a survey of 110 LT centers revealing it to be the preferred strategy in 50% to 60%. 13 However, it predisposes the other 50% of patients to receiving unnecessary prophylaxis which subjects them to an unnecessary risk of side effects (15%-22% of patients experience neutropenia ), potential for antiviral resistance, and the risk of late CMV after cessation of prophylaxis. The high risk of late CMV is highlighted by the fact that 57% of D1/R recipients still had a QFN-CMV < 0.1 IU/mL after 6 months of

5 1482 SOOD ET AL. LIVER TRANSPLANTATION, December 2015 Figure 4. Outcome to 6 months after transplant based on (A) QFN-CMV < 0.1 IU/mL and (B) QFN-CMV < 0.2 IU/mL. Figure 5. Kaplan-Meier curves of CMV DNAemia-free survival in 38 R1 patients based on week 2 results. CMV DNAemia occurred in 19 patients, compared with 19 patients who remained CMV-free. (A) Week 2 QFN-CMV < 0.1 IU/mL; (B) week 2 QFN-CMV < 0.2 IU/mL; and (C) week 2 QFN-MIT < 0.1 IU/mL. prophylaxis suggesting the provision of prophylaxis impairs the immune system s ability to form a sufficient immunoprotective response toward CMV. The alternate strategy, which is employed at our institution, involves preemptive management in which low-risk patients are monitored for a rise in their CMV PCR viral load. This strategy presents logistical issues, with ongoing extensive blood PCR monitoring and its associated costs. Furthermore, there is concern that patients are offered prophylaxis after they have already developed detectable CMV. Both strategies are effective, with a meta-analysis demonstrating a reduction of 80% in the incidence of CMV disease with universal prophylaxis and 60% using preemptive management. 17 However, improved individual assessment of immune responses toward CMV in the form of QFN-CMV could offer substantial improvements to risk stratification. In this study, we show that a single time point measurement of QFN-CMV at 2 weeks after transplant is significantly associated with development of DNAemia up to 6 months after transplant (Fig. 5A; HR, 6.9). We chose a primary cutoff of 0.1 IU/mL instead of 0.2 IU/mL as recent studies have determined 0.1 IU/mL a more suitable threshold for the QFN-CMV assay in

6 LIVER TRANSPLANTATION, Vol. 21, No. 12, 2015 SOOD ET AL Figure 6. Proposed protocol for management of CMV based on individualized immune response. transplant recipients. 11 Increasing the cutoff to 0.2 IU/mL improved sensitivity but reduced the specificity. Figure 4 shows that both cutoffs still result in an improved stratification of patients, given that 80% of patients with a cutoff of <0.1 IU/mL will develop CMV DNAemia despite being currently labeled low risk based on serology. The week 2 blood test was designated as the most appropriate and clinically useful time point based on several factors. First, after the initial nadir in QFN- CMV values at approximately day 3, there was a recovery to a new baseline of QFN-CMV between week 2 and month 1, allowing improved assay specificity. Second, episodes of detectable CMV DNAemia occurred well after the week 2 blood test with the viral load threshold being detected and treated at a median of 34 days after transplant. Any clinically useful immunological marker would need to be performed with results available in time to commence prophylaxis. The current manufacturer s recommendations are that the QFN-CMV result is dependent on a positive QFN-MIT titer. This is based on data for the QFN-gold assay which are often used in immunocompetent individuals. PHA MIT is a nonspecific antigen and is likely an inappropriate positive control (whether a positive control is necessary at all), especially in an immunosuppressed transplant population where overall immune response is diminished. As such, it is not unexpected that low QFN-MIT is associated with future risk of CMV. Similar findings have been suggested in a study of 41 bone marrow transplant recipients where early QFN- MIT nonreactivity was associated with risk of CMV reactivation. 18 Rather than being indeterminate, patients with low QFN-MIT are high risk for CMV and should begin prophylaxis. However, a nonreactive QFN-MIT has poor sensitivity and would not provide significantly more information than the QFN-CMV tube. It would therefore seem to be an unnecessary complication to any management algorithm. Our institution s threshold for treating CMV is relatively low at 1000 copies/ml. However, there are few natural history studies available to assist in defining the trigger points for intervention in a preemptive strategy. 9 Prior studies investigating preemptive strategies have employed a wide variety of possible plasma cutoffs which may have been impacted by methodology and range from 400 to 10,000 copies/ml. 19 As such, it is possible that some of the low-risk patients treated in this study may not have needed treatment and may have had spontaneous reductions of their CMV titers without any adverse effects. Another possible limitation of this study is that we did not include follow-up until 12 months, and it may therefore miss some cases of late CMV. Late CMV usually occurs after 6 months in recipients who had already been on prophylaxis or treatment, whereas our primary focus in this study was to improve risk stratification and identify patients who require early prophylaxis. Recent international consensus guidelines on the management of CMV following solid organ transplantation make mention that for clinical application an assay should ideally be simple to perform, inexpensive, highly reproducible, and amenable to either widely available platforms or shipping to specialized reference laboratories. 9 Despite the international guidelines, a survey suggested only 12% of units employed a form of CMVspecific T cell monitoring; however, the survey s authors conclude that it may in the future provide additional evidence to help guide decisions. 20 QFN- CMV is simple to perform and based on an already widely available laboratory platform. It has at least

7 1484 SOOD ET AL. LIVER TRANSPLANTATION, December 2015 equivalent sensitivity for CMV epitopes than enzymelinked immunosorbent spot (ELISPOT), 10 and despite only 22 haplotypes represented in the test, only <2% of the population have some rare human leukocyte antigen haplotypes that may result in false-negative results. 21 The assay does risk a higher false-negative rate as it only evaluates CD8 T cell responses, whereas studies have shown that long-term antiviral control of CMV is also mediated by CD4 T cells. 22,23 To date, QFN-CMV has not been well studied in LT or to assess risk of CMV as part of preemptive strategies. The largest and most comprehensive studies investigating QFN-CMV by Manuel et al. 11 and Kumar et al. 24 have focused on risk of late CMV after prophylaxis has ceased rather than on detecting which patients should have CMV prophylaxis commenced in the first place. These studies involved multiple transplant populations, with LT only comprising a minority of recipients (17%-21%). A study by Cantisan et al. 25 instead focused on pretransplant QFN-CMV, but in lung and kidney organ recipients. Although statistically significant, only 7 of 14 (50.0%) R1 patients with a nonreactive pretransplant QFN-CMV developed CMV, compared with 4 of 30 (13.3%) patients with a reactive result (P ). 25 Another study by Abate et al. 26 was part of a preemptive approach in kidney transplantation. However, this study was designed to compare ELISPOT with QFN-CMV, and although suggesting an association of QFN-CMV with CMV DNAemia, did not provide a clear description of time points most valuable to test at. 26 We propose a new protocol for management of CMV in low-risk transplant recipients who could be restratified with a simple once-off individualized blood test (Fig. 6). Patients would be restratified from low risk (R1) into high, moderate, or low risk. Low and moderate risk patients could have appropriate monitoring, whereas high-risk patients could commence prophylaxis. This protocol may be improved by further repeating QFN-CMV 6 months after prophylaxis to determine whether it should be continued longer term, given results of prior studies. 11 This proposed protocol would need validation in a prospective randomized controlled trial, but we believe it would reduce unnecessary use of antiviral medication (and associated side effects) in units employing universal prophylaxis while offering prophylaxis to patients who have been wrongly stratified as low risk in preemptive protocols. CMV reactivation is the most common infection following transplantation. Ideal management remains controversial. In this prospective observational cohort study, we show that early QFN-CMV performed 2 weeks after transplant could be used to improve risk stratification profiles of those patients currently labeled as low risk for development of CMV DNAemia. This individualization and optimization of patient management would be a significant advance in the care of LT recipients that could offer significant benefits to both patients and transplant units. REFERENCES 1. Lochmanova A, Lochman I, Tomaskova H, Marsalkova P, Raszka J, Mrazek J, et al. Quantiferon-CMV test in prediction of cytomegalovirus infection after kidney transplantation. Transplant Proc 2010;42: Sester M, Sester U, G artner BC, Girndt M, Meyerhans A, K ohler H. Dominance of virus-specific CD8 T cells in human primary cytomegalovirus infection. J Am Soc Nephrol 2002;13: Gandhi MK, Khanna R. Human cytomegalovirus: clinical aspects, immune regulation, and emerging treatments. Lancet Infect Dis 2004;4: Alford CA, Stagno S, Pass RF, Britt WJ. Congenital and perinatal cytomegalovirus infections. Rev Infect Dis 1990; 12(suppl 7):S745-S Boeckh M, Nichols WG. Immunosuppressive effects of beta-herpesviruses. Herpes 2003;10: Bosch W, Heckman MG, Diehl NN, Shalev JA, Pungpapong S, Hellinger WC. Association of cytomegalovirus infection and disease with death and graft loss after liver transplant in high-risk recipients. Am J Transplant 2011;11: Razonable RR. Cytomegalovirus infection after liver transplantation: current concepts and challenges. World J Gastroenterol 2008;14: Lautenschlager I. CMV infection, diagnosis and antiviral strategies after liver transplantation. Transpl Int 2009; 22: Kotton CN, Kumar D, Caliendo AM, Asberg A, Chou S, Danziger-Isakov L, Humar A; for Transplantation Society International CMV Consensus Group. Updated international consensus guidelines on the management of cytomegalovirus in solid-organ transplantation. Transplantation 2013; 96: Walker S, Fazou C, Crough T, Holdsworth R, Kiely P, Veale M, et al. Ex vivo monitoring of human cytomegalovirusspecific CD81 T-cell responses using QuantiFERON-CMV. Transpl Infect Dis 2007;9: Manuel O, Husain S, Kumar D, Zayas C, Mawhorter S, Levi ME, et al. Assessment of cytomegalovirus-specific cell-mediated immunity for the prediction of cytomegalovirus disease in high-risk solid-organ transplant recipients: a multicenter cohort study. Clin Infect Dis 2013; 56: Sanchez JL, Storch GA. Multiplex, quantitative, real-time PCR assay for cytomegalovirus and human DNA. J Clin Microbiol 2002;40: Levitsky J, Singh N, Wagener MM, Stosor V, Abecassis M, Ison MG. A survey of CMV prevention strategies after liver transplantation. Am J Transplant 2008;8: Manuel O, Venetz JP, Fellay J, Wasserfallen JB, Sturzenegger N, Fontana M, et al. Efficacy and safety of universal valganciclovir prophylaxis combined with a tacrolimus/mycophenolate-based regimen in kidney transplantation. Swiss Med Wkly 2007;137: Humar A, Lebranchu Y, Vincenti F, Blumberg EA, Punch JD, Limaye AP, et al. The efficacy and safety of 200 days valganciclovir cytomegalovirus prophylaxis in high-risk kidney transplant recipients. Am J Transplant 2010;10: Martın-Gandul C, Perez-Romero P, Gonzalez-Roncero FM, Berdaguer S, Gomez MA, Lage E, et al.; for Spanish Network for Research in Infectious Diseases REIPI. Clinical impact of neutropenia related with the preemptive therapy of CMV infection in solid organ transplant recipients. J Infect 2014;69: Kalil AC, Levitsky J, Lyden E, Stoner J, Freifeld AG. Meta-analysis: the efficacy of strategies to prevent organ

8 LIVER TRANSPLANTATION, Vol. 21, No. 12, 2015 SOOD ET AL disease by cytomegalovirus in solid organ transplant recipients. Ann Intern Med 2005;143: Tey SK, Davenport MP, Hill GR, Kennedy GA, Durrant ST, Khanna R, Cromer D, et al. Post transplant CMVspecific T-cell immune reconstitution in the absence of global T-cell immunity is associated with a high risk of subsequent virus reactivation. Bone Marrow Transplant 2015;50: Kandiel A, Lashner B. Cytomegalovirus colitis complicating inflammatory bowel disease. Am J Gastroenterol 2006;101: Le Page AK, Jager MM, Kotton CN, Simoons-Smit A, Rawlinson WD. International survey of cytomegalovirus management in solid organ transplantation after the publication of consensus guidelines. Transplantation 2013;95: Giulieri S, Manuel O. QuantiFERONVR -CMV assay for the assessment of cytomegalovirus cell-mediated immunity. Expert Rev Mol Diagn 2011;11: Sester M, Sester U, G artner B, Heine G, Girndt M, Mueller-Lantzsch N, et al. Levels of virus-specific CD4 T cells correlate with cytomegalovirus control and predict virus-induced disease after renal transplantation. Transplantation 2001;71: Walker JD, Maier CL, Pober JS. Cytomegalovirus-infected human endothelial cells can stimulate allogeneic CD41 memory T cells by releasing antigenic exosomes. J Immunol 2009;182: Kumar D, Chernenko S, Moussa G, Cobos I, Manuel O, Preiksaitis J, et al. Cell-mediated immunity to predict cytomegalovirus disease in high-risk solid organ transplant recipients. Am J Transplant 2009;9: Cantisan S, Lara R, Montejo M, Redel J, Rodrıguez-Benot A, Gutierrez-Aroca J, et al. Pretransplant interferon-g secretion by CMV-specific CD81 T cells informs the risk of CMV replication after transplantation. Am J Transplant 2013;13: Abate D, Saldan A, Mengoli C, Fiscon M, Silvestre C, Fallico L, et al. Comparison of cytomegalovirus (CMV) enzyme-linked immunosorbent spot and CMV quantiferon gamma interferon-releasing assays in assessing risk of CMV infection in kidney transplant recipients. J Clin Microbiol 2013;51: