Over the last few years there has been an impressive increase in the virological and immunological tools available to detect both human herpesvirus (HHV) and immune control of replication post-solid organ transplantation. This has allowed a greater appreciation of pathogenesis, studies to be designed to evaluate potential vaccines, new approaches adopted for antiviral deployment and the success of interventions to be judged. This chapter aims to summarize the state-of-the-art in vaccine development and look forward to the role that vaccines, immune monitoring, viral kinetics and new antiherpesvirus agents may play in the future management of HHV infections after transplantation.
The following text aims to illustrate the ways in which modern molecular and immunological technology has impacted on our understanding of HHV biology posttransplantation and has opened up new areas of opportunity especially in vaccine development and patient management. It will become apparent that HHV research post-solid organ transplantation has evolved to be based on infections that have historically been associated with a significant disease burden after SOT (CMV and to a lesser extent EBV) whereas others such as HSV or HHV 6, 7 and 8 can either be controlled through effective prophylactic therapy (HSV) or are not associated with significant problems after SOT (HHV-6, 7, 8). In order to put these studies in context, allusions to the situation in SCT recipients will be made where appropriate.
Vaccination to Prevent Human Herpesvirus Infections After Transplantation
Although there has been a relatively long period of inactivity in the CMV vaccine area after the ground breaking studies of Plotkin and colleagues in the 1970/1980s using the live attenuated Towne strain of CMV [1, 2], the last 5 years have seen some impressive advances. One of the challenges has been to identify the type of immune responses that a prototype vaccine geared toward the transplant recipient needs to generate. There is a significant body of data indicating that T cell responses against CMV especially their quality in the context of polyfunctional cytokine production are key to the control of CMV after both SCT and SOT [3-8]. However, it is known that antibodies against one of the major CMV glycoproteins, glycoprotein B (gB), can protect against vertical transmission in animal model systems  and the recent identification of a pentameric complex containing gH, gL, UL128, UL30 and UL131 as a major entry component and neutralizing antibody target [10, 11] provides an alternative strategy to vaccine development rather than solely relying on the generation of T cell responses. It is also worth bearing in mind that the immunosuppressive drugs given posttransplant have profound effects on T cell differentiation and function , particularly CD4 T cells, and so the ability of a vaccine to produce a good B and/or T cell response in a healthy immunocompetent individual may be reduced in the immuncompromised host with end-stage organ disease.
Notwithstanding the above issues, a recombinant gB vaccine administered with the potent adjuvant MF59 is the most advanced CMV vaccine and has been subjected to a number of phase 1 studies where it has demonstrated excellent immunogenicity and safety in seronegative individuals and can boost immunity in previously CMV exposed individuals (reviewed in ) This vaccine is a recombinant soluble form of gB in which the transmembraneous sequence has been removed together with mutation of the extracellular protease cleavage site and is produced in large-scale CHO cell cultures. At present, two phase 2 studies, one in seronegative women and one in solid organ transplant recipients (liver and renal patients) have been performed [13, 14]. The results from the phase 2 study in women of child-bearing age with existing children in playgroups, i.e. women who would be at high risk of experiencing primary CMV infection, indicated that the vaccine was 50% efficacious at preventing CMV infection . This appeared to be mainly driven through the induction of B cell immunity through IgG responses since only a transient increase in CD4 T cell responses could be detected . Recently, the results of a phase 2 trial in liver and renal transplant patients awaiting transplantation have been reported . In these patients the vaccine was highly immunogenic and well tolerated giving rise to high levels of anti-gB antibodies although relatively low titer neutralizing antibodies appeared to be induced using the standard neutralization assays. The titers of anti-gB antibodies generated in seronegative vaccines were similar to those observed in naturally seropositive individuals. In addition, a significant boost in the titer of anti-gB antibodies was observed in CMV seropositive vaccine recipients. Hence, despite these patients suffering from advanced end-stage renal and liver failure their ability to mount an antibody response against the gB—MF59 vaccine was very impressive. Perhaps the most interesting aspect of this study was the effect that vaccination had on posttransplant CMV infection. Two end-points are worthy of note both of which have clinical relevance. The first relates to the incidence of infection. In the high risk D+R− group, vaccine recipients had a lower incidence of CMV infection compared to the placebo group. This result was somewhat unexpected since the assumption would be that T cell immune responses would be a major protective feature interrupting transmission in such individuals. The precise mechanisms underlying this apparent protection are still being elucidated but may relate to anti-gB antibodies being involved in ADCC in addition to their ability to neutralize infectivity. A second parameter was also used to assess the effectiveness of the gB vaccine in the transplant setting whereby the duration of antiviral therapy needed to control replication was assessed. CMV seronegative recipients who received a donor seropositive organ and who had been vaccinated showed a significant reduction in the durations of DNAemia (p = 0.048) and the number of days of antiviral therapy required (p = 0.0287) in a preemptive management setting. There was a strong relationship between the titer of anti-gB antibodies at the time of transplant and the duration of viremia after transplantation in the entire population (p = 0.0022). These results are important for three reasons. Firstly, they demonstrate that generation of anti-gB responses to recombinant CMV antigens is a plausible approach in patients with end-stage liver and renal failure; second they indicate that anti-gB antibodies alone can offer some protection against infection and high level replication posttransplantation and third they reinforce the view that the transplant population is unique patient cohort in which to evaluate prototype vaccines especially when combined with pharmacodynamic readouts.
These results coupled with those recently described for a DNA vaccine encoding gB and the major T cell target pp65 antigen in stem cell transplant patients  pave the way forward for CMV vaccine trials in transplantation. In addition to the DNA vaccine and recombinant gB vaccine other approaches are being evaluated including a recombinant alphavirus replicon which expresses gB and an pp65-IE1 fusion protein and which has shown impressive immunogenicity in phase 1 trials . This vaccine-induced CMV-specific CD8 T cells with a IFNγ and TNFα polyfunctional phenotype and CD4 T cells that were able to produce both IFNγ and IL-2—a phenotype that has been associated with protection against high level replication and disease in other studies [3-8]. However, outside of the DNA vaccine and recombinant gB vaccine these other prototype vaccines have yet to be deployed in the transplant setting. If there was a vaccine for CMV available then one could envisage it being deployed in seronegative patients prior to transplantation since these are at greatest risk of serious CMV disease posttransplant or potentially toward the end of a prophylactic period posttransplant to minimize postprophylactic CMV syndrome/disease. However, boosting immunity in a CMV seropositive recipient could also be valuable as there is still a significant disease burden in this group despite the availability of prophylactic and preemptive therapy.
A summary of the approaches being deployed to develop a CMV vaccine and their ability to generate B and T cell immune responses are shown in Figure 1.
Varicella zoster vaccines
While the majority of adult transplant patients are already exposed to VZV (∼90%) there are some that remain at risk because they are seronegative and in the context of the pediatric population a larger proportion may not have been exposed to VZV. The severity of posttransplant VZV infection and disease should not be underestimated especially herpes zoster which has an incidence of approximately 10% in the first 4 years post-SOT [18, 19]. As vaccine uptake with the Oka strain increases it is likely that that the proportion of SOT patients at risk will decline in the future but globally not all countries have adopted a comprehensive VZV immunization policy. VZV vaccination with the Oka must be performed pretransplant. The vaccine is effective in patients with end-stage renal and liver disease although seroconversion rates appear to be lower in these patents compared to healthy individuals. Vaccination with two doses is recommended 4–6 weeks prior to transplantation. Although the use of live attenuated vaccines post-SOT is contraindicated, the Oka strain has been used in this context but further controlled trials must be performed before this approach can be widely deployed (reviewed in ). The herpes zoster vaccine (Zostavax, Merck and Co.) contains a substantially increased infectious dose of live virus and is therefore not recommended posttransplantation and further studies in patients with end-stage renal and liver disease are awaited. However, in the future such a vaccine may prove effective in minimizing the risk of herpes zoster infections and their inherent complications after SOT.
Progress toward an EBV vaccine has been less impressive compared to CMV. Most studies have focused on the major viral surface glycoprotein gp350 which is known to induce neutralizing antibodies. A study in seronegative healthy volunteers of a recombinant form of gp350 administered with an alum/monophosphoryl lipid A adjuvant did not protect against EBV infection but did reduce the incidence of infectious mononucleosis by 78% . The same vaccine but with alum as an adjuvant has been evaluated in small number of renal dialysis patients and showed relatively poor immunogenicity . A hybrid peptide-based vaccine comprising a segment of the EBNA-3 coupled with tetanus toxoid has been evaluated in healthy volunteers and produced T cell responses  but has yet to be evaluated in patients awaiting transplantation.
Through the development of acyclovir, valaciclovir and related molecules, the impact of HSV infections after transplantation has been minimized substantially. However, there remains attractiveness to developing an HSV vaccine to either prevent primary infection or as a prophylactic vaccine. A number of vaccine candidates have been evaluated in the clinic or are undergoing development (reviewed in ). Over the years the development of an HSV vaccine has been a story of mixed fortunes. In the context of the solid organ transplant recipient the safety features associated with replication incompetent vaccine preparations are attractive and vaccines based on viral glycoproteins such as glycoprotein D delivered through a DNA vaccine or alphavirus replicon system are in the early stages of development . Subunit vaccines based upon gD have proven generally disappointing although some evidence of efficacy has been noted in HSV-2 seronegative women (reviewed in . It is the view of the author that it is unlikely that a vaccine candidate for HSV will be subjected to evaluation in the transplant setting by 2017.
Use of Viral Kinetics to Personalize Therapy for HHV Infections
The ability to accurately determine viral load for HHV infections posttransplantation has become a mainstay for diagnostics especially in the context of the beta and gamma herpesviruses. Most approaches use real-time quantitative PCR-based assays. Current clinical management protocols frequently use viral load threshold to initiate therapy for CMV when operating in preemptive therapy mode and for monitoring viral load patterns subsequent to the initiation of therapy [25-28]. Initial transient rises or a static viral load in the early stages of therapy are not uncommon and reflect the natural dynamics of CMV infection in vivo . However, persistent viral replication in the absence of other factors such as noncompliance may indicate drug resistance . Despite the widespread use of quantitative measures for CMV over many years, a consensus on when to initiate therapy is not available. This situation has been exacerbated by the variation between different assays and different laboratories making comparisons difficult. The recent availability of an international standard for CMV genome quantification (provided by the National Institute of Biological standards and Control (NIBSC) UK) should impact positively in this area and in the future will facilitate multicenter comparisons and trials of new antivirals and vaccines. Ultimately, a consensus on the optimal CMV load to initiate preemptive therapy which both minimizes overtreatment while preventing the majority of CMV syndrome and disease should be developed. While this would be of value in a variety of SOT and SCT settings, the ability to be able to integrate viral load kinetics and other risk factors such as T and B cell immune responses and host genetic factors should not be overlooked .
What does the future hold in the context of CMV replication kinetics? We are now seeing pharmacokinetics being incorporated into trials of new anti-CMV compounds and vaccines allowing both dose optimization to occur prior to performing larger phase 3 trials and surrogate end-points for efficacy to be documented [14, 16, 32]. With the availability of increasingly complex mathematical models it is likely that the predictive nature of viral kinetics, prior to or shortly after antiviral therapy, may allow the development of an individualized approach to therapy especially when the genetic composition of CMV is included additionally in such algorithms [33-35]. The availability of next-generation sequencing methodologies make this combination particularly powerful and able to be delivered in real time. Knowledge of viral replication kinetics in different clinical settings (transplant group, induction immunosuppression, augmented immunosuppression) allows us to determine directly the basic reproductive number of HHV infections [24, 36, 37]. This is most well developed for CMV and provides the baseline data against which the efficacy levels needed to control infection after SOT within these different groups can be estimated and the success, or otherwise, of both antiviral or vaccine preparations assessed accordingly.
Outside of the work on CMV replication kinetics, the other HHVs have not been subject to such intense investigation. EBV viral load monitoring is now routine and high viral loads are often associated with concurrent PTLD but data relating EBV kinetics to the risk of developing PTLD remain controversial . Similar to the situation with CMV, the lack of an international genome standard for quantification of EBV in molecular assays makes comparison of thresholds for impending PTLD difficult to interpret. Approaches using the cumulative EBV load experienced by a patient and the combination of EBV load measures with T cell immunity are now being reported. Estimates of R0 for EBV post-SCT have been provided and the effects of rituximab on EBV DNAemia have been quantified in one study . It is likely that these approaches will facilitate the management of EBV post-SOT in the future.
Use of Immune Markers to Identify “At Risk” Patients
Recent years have seen an increasingly availability of more sophisticated immunologic reagents to determine both T cell frequency and function. Both CD8 and CD4 T cells are important in the control of CMV replication in healthy individuals and in the posttransplant period. The deployment of class I HLA multimers has enabled a new appreciation of the CD8 T cell responses against HHV infections in both immunocompetent and the immunocompromised host (reviewed in ). Multicolor flow cytometry has also expanded the breadth of CD4 and CD8 T cell responses that can be enumerated and so polyfunctional T cell responses can now be determined relatively easily. These advances are now impacting in the transplant setting and it is likely that further deployment of these approaches will lead to the routine immunophenotyping of transplant patients in the near future . The utility of simpler assays such as the Quantiferon-IFNy assay argues that routine deployment of such assays is likely [41, 42]. The major question is not whether such assays provide insight into the immune control of infections (which are essential for the development of vaccines) and identifying immune control phenotypes (e.g. PD-1 expression) but whether they add significant value in risk stratification of patients with respect to their risk of future HHV infections. The most persuasive evidence in this context derives from CMV postsolid organ transplantation where the functional capacity of CMV specific CD8 and CD4 T cells to produce IFNy and poyfunctional cytokines has been shown to be predictive of the risk of future high level CMV DNAemia and CMV disease [6, 7, 43]. While such approaches are not routinely deployed at present, many centers, including our own, are investigating complementing routine viral load measurements with routine T cell surveillance to identify patients who are at high risk of CMV complications. The identification of patients who are at risk of late CMV infection and disease after a period of prophylaxis remains an important goal in the SOT setting. Recent data suggest that the CMV T cell responses at the end of prophylaxis may allow risk stratification of patients . In the study by Kumar et al., 35% of patients had a CD8 T cell response against CMV using the Quanitferon assay and late CMV disease occurred in 5.3% of these patients compared to 23% of those who did not develop a CD8 T cell responses. In the high risk D+R− population patients exhibiting a CD8 T cell response at day 100 had an incidence of CMV disease of 10% whereas 40% of patients with no evidence of a CD8 T cell response against CMV progressed to CMV disease.
Outside of CMV, less information is available for the routine monitoring of T cell responses against other HHV infections. Multiple reagents are available for enumerating EBV T cell responses although relatively few studies have addressed the risk for PTLD in the context of pathogen specific T cell responses. One of the reasons why such reagents are not deployed against other HHV infections after SOT is that relatively few immunodominant T cell targets and individual epitopes have been identified in these other HHV pathogens.
Adoptive Immunotherapeutic Developments
Adoptive immunotherapy has tended to be deployed in patients poststem cell transplantation. Indeed, much of the early ground breaking work was performed with CMV T cells by the Seattle group (reviewed in ). The use of adoptive T cell therapy to control CMV infection post-SCT is an area of active interest with two clinical trials in progress and a number of others in the planning stages. However, in the solid organ transplant setting adoptive immunotherapy for CMV has not been pursued aggressively since immunosuppression is not as dramatic as that observed post-SCT, i.e. the residual immune system still retains functionality after SOT which contributes to CMV control and the availability of CMV prophylactic drugs such as valganciclovir minimizes CMV disease risks further . In contrast, EBV infection and in particular EBV-driven PTLD is a more difficult disease to manage with little evidence that antiherpes drugs are effective especially once PTLD is manifest (reviewed in ). Guidelines are available for the management of PTLD posttransplantation (e.g. ). An alternative approach is to generate T cells for adoptive immunotherapy of PLTD. One controlled study has been reported with encouraging results [48, 49]. Haque et al. used a donor lymphocyte bank to expand EBV-specific T cells which were then available to patients who were experiencing PTLD postsolid organ transplantation. CTL clones were chosen on the basis of their maximal killing of EBV+ LCLs from PTLD patients combined with minimal killing of EBV negative blast cells from the same patient. In a controlled study of 28 patients with PTLD from a variety of solid organ transplant settings a complete response (using tumor regression, improved graft function and decreases in EBV load as markers) was achieved in 14/28 patients and a partial response observed in a further three patients . Long-term mortality at 2 years post-CTL therapy was 21%. Interestingly, better responses were observed in female patients, those with early onset PTLD after transplantation, patients with a single site tumor where tumor burden was relatively low and in patients with primary rather than reactivated EBV infection . Clinical grade VZV-specific T cells for adoptive immunotherapy in SCT patients have recently been described which may also prove useful in the SOT setting in difficult to manage cases of VZV .
New Antivirals for Control of HHV Infections
The promise of Maribavir (Figure 2) for the prophylactic control of CMV was proven to be unfounded in phase 3 clinical trials in SCT patients  and the phase 3 trial in liver transplant patients was terminated early on the grounds of lack of efficacy. The disappointment of the Maribavir trials should be viewed in the context of the study design (for example, patients were given the drug only at engraftment in the SCT trial and CMV disease was used as an endpoint) and whether the optimal dose was deployed (the dose was based on a phase 2 study and was probably at the lower end of the efficacy window). These factors will inform the design of future trials of antivirals against HHV either as prophylactic or preemptive agents. The use of pharmacokinetics to determine the appropriate dose is becoming standard practice and will allow a dose of drug to be chosen that has the correct balance of high efficacy and low toxicity. In the context of CMV, AIC246 is a specific inhibitor of the UL56 terminase complex  and has been evaluated in phase 1 and recently in phase 2 studies (www.aicuris.com). The drug, now named Letermovir, is well tolerated and provides rapid control of HCMV replication in patients with DNAemia. In the phase 1 study, 85.7% and 88.9% of patients receiving a dose of AIC246 of 40 mg bid (n = 7) or 80 mg QD (n = 9) respectively cleared HCMV DNAemia in whole blood over the 14 days of therapy compared to 55.6% in the observational control group (n = 9). Similar effects were observed for HCMV plasma DNAemia with 71.4% clearance by day 14 in the 40 mg bid arm and 55.6% in the 80 mg QD arm. The drug has also been used to treat established HCMV disease using doses of 120 mg and 240 mg QD in small numbers of patients and for treating patients with GCV resistant and multidrug resistant HCMV infection. In a phase 2b dose ranging study (120 mg od and 240 mg od vs. placebo) in 133 HCMV seropositive SCT patients both doses of Letermovir showed a significant reduction in prophylaxis failure in the study drug arms compared to placebo and in the high-dose group (240 mg od) the time to prophylaxis failure was significantly longer (www.aicuris.com). More extensive trials of this drug are planned and it will be a potent addition to the armamentarium against CMV not least because it targets a unique aspect of the CMV life cycle which complements existing DNA polymerase-based inhibitors.
An equally interesting compound which has the potential to target a broad range of herpesvirus infections (and other nonherpes infections that are problematic for the transplant recipient such as BK virus and adenoviruses) is the lipid conjugated form of cidofovir, CMX001 . The parent drug cidofovir is a potent inhibitor of the HHV DNA polymerases but its poor oral bioavailability and predilection for concentration in the proximal tubules of the kidney limits its use. In contrast, the haxadecyloxypropyl ester (CMX001; Figure 2) demonstrates greatly enhanced bioavailability with an IC50 for CMV of 0.9 nM compared to 0.38 uM for the parent cidofovir . CMX001 has undergone favorable phase I clinical studies and has been evaluated for safety, tolerability and efficacy at preventing CMV in a phase 2 dose escalation study in 230 CMV seropositive stem cell transplant recipients (www.chimerix.com). The results indicate that CMX001 at a dose of 100 mg twice a week led to a statistically significant decrease in the proportion of patients experiencing CMV DNAemia >200 genomes/mL or disease (p = 0.001 vs. placebo). The future prospects of this compound are very promising given its broad anti-HHV activities and reduced toxicity over cidofovir.
A further new inhibitor of CMV replication is the purine 2-(hydoxymethyl)methylenecyclopropane analog cyclopropavir which is a potent inhibitor of the HCMV DNA polymerase  and also inhibits UL97 function . Interestingly, the drug also inhibits HHV6 and HHV-8 replication. Cyclopropavir is approximately 10-fold more active against HCMV replication (EC50 = 0.4 uM) in vitro than GCV. Activation of cyclopropavir requires initial stereoselective phosphorylation by the UL97 kinase (or its homolog on HHV-6 and HHV-8) followed by cellular kinase phosphorylation to the active triphosphate  and so has the same drug activation characteristics as GCV. Mutations in the UL97 protein associated with cyclopropavir resistance overlap with, but are not identical to, GCV drug resistant kinase mutations. Mutations at M460I and H520Q conferred 12- and 20-fold increases in the EC50 whereas the mutations M460V, A594V, C592G and C603W were associated with a lower level of resistance (two- to fivefold). Interestingly, the L595S mutation associated with GCV resistance was fully sensitive to cyclopropavir while the novel mutations M460T and C603R conferred 8- to 10-fold increases in the EC50 . A valine ester of cyclopropavir has been synthesized  which leads to increased oral bioavailability (a similar modification was used to increase ganciclvoir bioavailablity) and paves the way forward for new clinical studies on this drug in the transplant setting.
There is no doubt that our understanding of HHV pathogenesis posttransplantation has improved dramatically over the last few years but challenges remain in optimizing patient management. There is a major opportunity for the development of a holistic approach to patient management which combines virologic, host immune responses and potentially viral and host genetics. New drugs are on the horizon for CMV and potentially a pan-herpesvirus drug in the form of the lipid conjugated derivative of cidofovir. At present, the evaluation of vaccines against CMV is a particularly highly dynamic area and is being extended to other HHV infections such as EBV and HSV. In all these infections one should expect to see the deployment of both established vaccine approaches and newer approaches including immune-based therapeutics which will ultimately lead to alternative options in preventing or minimizing the impact of these HHV infections, and particularly CMV infection, after transplantation.
The author has conflicts of interest to disclose as described by the American Journal of Transplantation. The author has received honoraria as a speaker and for advisory boards from Roche pharmaceuticals and Viropharma Inc.
Will there ever be a vaccine for cytomegalovirus in the immunocompromised host?
Only nonreplicating, subunit or peptide-based vaccines for human herpesvirus should be used in the posttransplant phase?
Will a pan-human herpesvirus antiviral drug ever become a reality?