Cytomegalovirus Infection in Solid Organ and Hematopoietic Cell Transplantation: State of the Evidence

Overview of Prophylaxis Versus Preemptive Therapy

The 2 strategies for the prevention of CMV in transplant recipients are antiviral prophylaxis and preemptive therapy, which are each considered the preferred prevention approaches after SOT and HCT, respectively (Table 1). Each strategy has its own specific advantages and disadvantages. Prophylaxis after SOT has been shown to reduce not only CMV infection and disease while patients are receiving antiviral therapy, but also graft loss, mortality, and opportunistic infections [2]. All CMV D⁺/R^– or R⁺ SOT recipients being managed with prophylaxis typically receive oral valganciclovir for a defined duration, usually 3–6 months, but up to 12 months after lung transplantation [1, 3, 4]. In contrast, the overall benefit of prophylaxis after HCT has been difficult to assess [5]. Thus, HCT recipients have traditionally been managed with preemptive therapy, whereby they are monitored for the development of asymptomatic CMV replication and are only given antivirals upon the detection of viremia, with the goal of preventing progression to disease. The long-term benefits of preemptive therapy are postulated to occur as a result of the development of CMV-specific immunity due to controlled viremia [1, 2]. As discussed below, however, there are emerging data demonstrating the benefits of preemptive therapy even after SOT, though this approach is not currently standard of care at most transplant centers, in part due to logistical considerations. A hybrid strategy whereby initial antiviral prophylaxis is followed by surveillance for CMV viremia with preemptive therapy is often used in SOT recipients [6], though its benefit remains unproven. To date, the optimal approach for the prevention of CMV remains a controversial issue.

Prophylaxis, Delayed-Onset CMV Disease, and Neutropenia

Prophylaxis after SOT prevents CMV infection during the period when patients are receiving antivirals [3, 4], but is associated with high rates of postprophylaxis delayed-onset CMV disease, with resultant increased mortality during the first year after transplantation [7]. Extending the duration of prophylaxis, while adopted by many transplant centers, does not curtail the risk of postprophylaxis disease and leads to significant myelotoxicity [8, 9]. Many transplant centers monitor patients for the development of CMV viremia for a fixed duration upon discontinuation of prophylaxis. While intuitively appealing, this practice only detects a small proportion of patients with CMV infection, as most either develop infection after the monitoring period has ended, clear their viremia spontaneously, or progress to CMV disease prior to initiation of antivirals because of rapid viral load doubling [10]. Valganciclovir prophylaxis is also associated with high rates of neutropenia after SOT. In a study of 304 liver transplant recipients, neutropenia developed in 40% of patients receiving valganciclovir prophylaxis vs 16% of those not on prophylaxis and was an independent predictor of overall mortality [11]. To reduce the risk of neutropenia, many centers use low-dose valganciclovir prophylaxis, but this unfortunately leads to breakthrough viremia, particularly due to ganciclovir-resistant CMV [12].

A recent meta-analysis evaluated CMV prophylaxis in HCT recipients [5], although this approach remains infrequently used due to the myelotoxicity of ganciclovir/valganciclovir. In addition, while there are randomized trial data demonstrating the efficacy of acyclovir and valacyclovir for CMV prophylaxis after HCT [13–16], the use of these drugs is not widespread due to the need for intravenous administration of acyclovir and extremely high doses of valacyclovir, as well as the modest anti-CMV activity of these agents [17]. However, since recipient CMV seropositivity is associated with poor overall outcomes after HCT [18] and detectable plasma viremia is associated with overall mortality in seropositive HCT recipients [19], prophylaxis with a nontoxic drug or prevention with an effective vaccine may be beneficial.

Letermovir, a novel terminase complex inhibitor with no hematological toxicities, has been shown to reduce clinically significant CMV infection after HCT and is licensed for CMV prophylaxis among high-risk HCT recipients [20]. Limited clinical experience exists with the use of letermovir in the SOT setting, and resistance may develop earlier than with valganciclovir or foscarnet [21]. While it may not reduce the risk of postprophylaxis disease after SOT, it is not myelosuppressive and thus may be an option for patients with dose-limiting toxicity due to valganciclovir. A clinical trial of letermovir prophylaxis among CMV D⁺/R^– kidney transplant recipients is under way (ClinicalTrials.gov identifier NCT03443869) [16].

Preemptive Therapy

CMV viremia is inherent with preemptive therapy but occurs in a carefully controlled manner, which allows for the rapid initiation of antivirals. The major benefit of preemptive therapy compared to prophylaxis after SOT is the lower frequency of delayed-onset CMV disease [22, 23]. The biological basis of this finding is that controlled CMV viremia simulates immunity as a natural CMV vaccine, resulting in greater antibody neutralization and CD8⁺ T-cell responses compared with prophylaxis [24]. Most trials comparing prophylaxis and preemptive therapy after SOT have focused on kidney transplant recipients and comprised a small number of D⁺/R^– patients [22, 25–27]. More recently, a randomized trial in 205 high-risk D⁺/R^– liver transplant recipients documented that preemptive therapy compared to prophylaxis for 100 days was efficacious and cost-effective for the prevention of CMV disease within 12 months posttransplant (NCT01552369) [28].

Nonetheless, published guidelines have generally recommended prophylaxis as the CMV prevention modality of choice after SOT [29], in part because of the large evidence base supporting the use of prophylaxis, its beneficial impact on the indirect effects of CMV, and the logistical difficulties of implementing a preemptive protocol at transplant centers. In addition, the frequency and duration of CMV monitoring, the viral load thresholds at which to start antivirals when viremia develops, and the optimal duration of antiviral therapy to treat controlled asymptomatic viremia have not been rigorously defined after SOT.

HCT recipients, in contrast, have been managed using preemptive therapy for years, allowing for the continuous optimization of this approach in HCT patients. Indeed, results in the standard-of-care preemptive therapy groups from recent randomized trials suggest excellent prevention of CMV disease by using this strategy (Table 2). However, while late monitoring after HCT is recommended for high-risk patients, adherence to weekly monitoring schedules is poor, and late disease continues to occur [19]. Nonetheless, in addition to the benefits of preemptive therapy in limiting drug toxicity and possibly accelerating immune reconstitution, the permissive CMV reactivation that occurs in this context may also have a protective effect on leukemic relapse [30, 31]. However, the issue remains controversial as the data are inconsistent, and the effect of CMV reactivation on nonrelapse mortality appears to outweigh any possible antileukemic effect. Larger analyses assessing the effect of preemptive therapy on all outcomes after HCT will be needed to answer this question.

Novel Approaches to Diagnose CMV Viremia

CMV surveillance in the late period after SOT and HCT has been advocated as part of extended monitoring or hybrid strategies of prophylaxis and preemptive therapy. However, adherence to the optimal weekly monitoring schedule has been a formidable obstacle due to logistical challenges in the late posttransplant period, preventing many patients from having regular office-based blood draws. The dried blood spot (DBS) technology has been used to diagnose CMV by polymerase chain reaction in the congenital CMV setting and has recently been adopted to design a self-testing platform, which accurately quantitates the CMV viral load in finger stick–collected DBSs [37]. A multicenter randomized controlled trial is currently ongoing in HCT recipients at high risk for late CMV complications to determine whether a mobile device–assisted home-based DBS self-testing method will result in higher adherence rates for late CMV surveillance (NCT03910478) [38].

Immune Monitoring

Monitoring of CMV-specific immunity has emerged as a promising new modality that may be able to shift the field of CMV prevention toward the paradigm of precision medicine. The largest study to date of the clinical utility of CMV immune monitoring comprised 537 kidney transplant recipients (260 D⁺/R^– and 277 R⁺) from 43 international centers, who were enrolled either pre- or posttransplantation and prospectively followed for 12 months after transplantation. Antiviral prophylaxis was administered as standard of care (generally 6 vs 3 months for the D⁺/R^– and R⁺ groups, respectively), and all patients had a T-SPOT.CMV assay checked within 5 days of completing antiviral prophylaxis and at 1, 2, 3, 4, and 6 months after prophylaxis. The assay measured interferon-γ release from CD4⁺ and CD8⁺ cells in response to CMV antigen stimulation. Among the 368 patients who were eligible for analysis, the presence of a positive cell-mediated immune assay at the end of prophylaxis predicted freedom from CMV events [39]. In contrast, patients with low levels of CMV cell-mediated immunity were more likely to experience CMV infection and disease. Additionally, pretransplant cell-mediated immunity was a predictor of future CMV events, whereas immunosuppression was not. However, these differences were most pronounced among CMV R⁺ kidney transplant recipients and not the “high risk” CMV D⁺/R^– patients, a minority of whom actually developed CMV-specific immunity by the end of antiviral prophylaxis. Thus, while the investigators were able to find a high negative predictive value for CMV infection among the few CMV D⁺/R^– who developed CMV-specific cell-mediated immunity, the performance characteristics of this test were overall extremely poor in this cohort. Nonetheless, these data suggest that such assays performed in both the pre- and posttransplant settings may be able to risk-stratify patients, particularly those who are CMV R⁺.

It remains to be determined how the use of CMV immune monitoring will be incorporated into clinical practice. Among the challenges that must be overcome are the lack of routine availability, high cost, slow turnaround time, and lack of standardization due to the multitude of assays available. Indeed, several commercial tests exist that measure interferon-γ production by CD4⁺ and/or CD8⁺ T cells in response to stimulation with CMV-specific antigens or peptides (eg, pp65 and intermediate-early 1 [IE-1] antigen) [6, 39, 40]. However, other markers may also accurately predict CMV-specific immunity, such as the T-box transcription factor (“T-bet”) or even complex polyfunctional T-cell and cytokine signatures [41, 42]. Furthermore, none of these platforms have incorporated the assessment of humoral immunity, which is important for the control of primary CMV infection and may also be able to predict postprophylaxis CMV disease risk [24]. In the current era, these technologies may be used, but clinicians will need to interpret their results with caution.

Monoclonal Antibodies

While the field of CMV prevention and immunity has largely focused on T-cell responses, neutralizing antibodies may also play a role in controlling CMV, particularly primary infection as previously mentioned. This has led to interest in the study of monoclonal antibodies as a method of CMV prevention. Unlike polyclonal antibodies (such as intravenous immunoglobulin), which were once commonly used for prophylaxis [43], monoclonal antibodies are more specific and less toxic, and may allow for the administration of higher doses of drug. A recent phase 2 trial of D⁺/R^– kidney transplant recipients demonstrated that RG7667, a combination of 2 anti-CMV monoclonal antibodies that inhibit entry of CMV into host cells, reduced the proportion of patients with CMV viremia within 24 weeks (but not 12 weeks) posttransplantation, significantly delayed the time to CMV viremia, and reduced the incidence of CMV disease [44]. Notably, although RG7667 did not eliminate the need for antivirals, the use of preemptive valganciclovir was lower in the RG7667 group compared to the placebo group, suggesting that its role may be as an adjunct to preemptive therapy among high-risk SOT recipients.

Vaccines

An ideal CMV vaccine should be able to prevent or modulate CMV replication and disease. Unfortunately, trials of CMV vaccines have yielded conflicting results [45]. A study from the 1980s of a live attenuated vaccine given to kidney transplant candidates found no reduction in the incidence of CMV infection, though infections were milder in patients who received the vaccine [46]. A proof-of-concept phase 2 trial of kidney and liver transplant candidates found that a CMV glycoprotein B (gB) vaccine with an MF59 adjuvant resulted in increased levels of gB antibody titers, with shorter duration of viremia and less ganciclovir requirement among vaccinated individuals [47]. In contrast, another phase 2 trial of a CMV plasmid vaccine including gB and pp65 did not prevent CMV infection in D⁺/R^– kidney transplant recipients, but this was thought to be due to its administration 30 days after transplantation [48]. A trial of a bivalent recombinant RNA viral vector subunit CMV vaccine that expresses pp65 and a truncated isoform of gB (dubbed HB101) against CMV infection among D⁺/R^– living-donor kidney transplant candidates (administered prior to transplant) is currently recruiting patients (NCT03629080) [45].

CMV vaccines have also been evaluated in seropositive HCT recipients. Preliminary results of a phase 3 trial using the gB/pp65 DNA vaccine showed a failure to prevent CMV disease or mortality [49]. A pp65 peptide vaccine showed promising results in a phase 1 trial [50], and recently reported results using a modified vaccinia Ankara peptide vaccine also showed a reduction in CMV viral loads in an early-phase clinical trial [51].

New Drugs for Refractory and Resistant CMV Infection

Ganciclovir-resistant CMV infection remains associated with poor outcomes, due in part to the significant toxicities of foscarnet and cidofovir [1, 52]. Similarly, outcomes of refractory CMV infection, defined as progressive CMV disease without the detection of UL97 or UL54 resistance mutations, remain suboptimal [1]. However, drug development for these indications has stalled. A phase 2 study of maribavir, a direct UL97 inhibitor, demonstrated efficacy against refractory or resistant CMV infections in transplant recipients [53], and a phase 3 trial is currently enrolling patients (NCT02931539) [50]. Unfortunately, there are currently no plans to further develop brincidofovir (an oral lipid conjugate form of cidofovir with no nephrotoxicity) due to its failure to meet the primary endpoints in prophylaxis studies of HCT recipients [1].

Until the ongoing maribavir trial is completed, few good options exist for the management of resistant and refractory CMV infection [54]. One currently available consideration is the off-label use of letermovir. This is an appealing drug because of its favorable side-effect profile, its availability in both oral and intravenous formulations, and its activity against CMV strains with UL97 and UL54 mutations. Unlike most other anti-CMV antivirals, it has multiple CYP system-mediated interactions [20]. However, reports of its use for the treatment (as opposed to prevention) of CMV infection in both SOT and HCT remain sparse, and the treatment dose remains undefined.

In a case series of 4 lung transplant recipients and 1 HCT recipient receiving letermovir, 4 of 5 patients experienced reductions in CMV viremia, but 1 patient experienced a rapid increase in viral load [55]. The latter treatment failure was thought to be due to the use of the lower 240-mg dose of letermovir as opposed to the standard 480-mg dose recommended for prophylaxis. Only 2 patients received letermovir monotherapy, with the rest receiving combination therapy with ganciclovir or foscarnet. Another case series of 2 lung and 2 heart transplant recipients with CMV retinitis showed that all patients had funduscopic improvement on letermovir, though 3 of them received concurrent intravitreal medications [56]. In addition, despite the fact that doses of 720–960 mg were used, 2 patients had persistent low-level viremia, and 1 had high-grade viremia. Most ominously, genotyping confirmed the presence of treatment-related UL56 resistance mutations, which have been shown to rapidly develop during serial viral passage experiments in vitro [57]. Clearly, large clinical trials evaluating the efficacy of dose-optimized letermovir, possibly in combination with other drugs, for the treatment of resistant and refractory CMV infection must be conducted.

Adoptive T-Cell Transfer

The use of adoptive CMV-specific T-cell therapy is a potential alternative to classic antiviral therapy for patients with refractory or resistant CMV infection. There are studies to support the use of this treatment modality among HCT recipients with viral infections [58, 59]. However, data in SOT recipients are sparse. In 2 CMV D⁺/R^– lung transplant recipients with ganciclovir-resistant CMV disease, adoptive T-cell therapy effectively cured the infection, provided long-term CMV-specific immunity, and resulted in no untoward consequences to the allograft [60, 61]. More recently, outcomes were reported for 13 SOT recipients (8 D⁺/R^–, 3 D⁺/R⁺, and 2 D^–/R^–) with drug-resistant or refractory CMV disease who prospectively received expanded autologous CMV-specific T cells [38]. Eleven of these 13 patients responded to therapy, with resolution or reduction of viremia, improvement in end-organ disease, and discontinuation or reduced use of antivirals. In addition, 4 patients had an increase in CMV-specific T-cell immunity, even after cessation of T-cell therapy.

There are challenges that must be overcome before such technologies can become sustainable for mainstream use, particularly the issues of limited donor availability (autologous vs allogeneic human leukocyte antigen–matched donors) and prolonged product generation time. The advent of third-party virus-specific T-cell banks for “off the shelf” administration of products is a promising new development in the field and will result in the on-demand availability of T-cell products [58]. This treatment modality will need to be explored in future studies.