The workshop “Cytomegalovirus Infection: Advancing Strategies for Prevention and Treatment,” sponsored by the National Institute of Allergy and Infectious Diseases and the Eunice Kennedy Shriver National Institute of Child Health and Human Development, was held in Rockville, Maryland, on 4–6 September 2018. The meeting objectives were (1) to review the current state of knowledge of cytomegalovirus (CMV) disease, including which populations may be most affected; (2) to assess challenges associated with current diagnostic platforms; and (3) to facilitate the discovery and development of vaccines and therapeutics for serious CMV infections and/or disease by summarizing the current state of knowledge and identifying possible barriers to product development. A brief summary of the major outcomes and identified research gaps are summarized here.
PATHOGENESIS
Much has been learned about the pathogenesis of CMV and how the viral immune evasion capabilities interplay with the host’s immune status and response. However, more data are needed to facilitate the development of effective preventative and therapeutic agents. CMV infections are usually benign in those with intact immune systems, although CMV may cause mononucleosis in some people. The virus is usually acquired mucosally. It can infect many types of cells, and its large genome encodes many genes that serve to evade the host immune response. In fact, gene content from ribosomal profiling has been shown to be much greater than predicted for a DNA virus, approaching the variability of some RNA viruses. Data presented suggested that a tumor-specific endogenous repetitive element, Human satellite II (HSATII), is induced in high-grade but not in low-grade CMV colitis. These observations support complex immune evasion mechanisms and suggest more of a role for human CMV in disease pathogenesis. However, CMV is most pathogenic in individuals who already have a degree of immunodeficiency. These manifestations include congenital infection, infection in transplant recipients either from reactivation or donation from a seropositive donor, and other forms of immunocompromise, such as those associated with chemotherapy.
Critically ill seropositive patients with conditions such as sepsis, burns, trauma, and cardiac comorbid conditions have been shown to have worse outcomes than their seronegative counterpoints. In addition, prior CMV infection and the establishment of latency have been implicated in several other diseases, although more research is needed to establish an exclusive or contributory causal relationship. The mechanism would presumably be based on low levels of viral gene expression during latency in ≥1 cell type. The conditions include immune senescence, heart disease, and vascular diseases such as arteriosclerosis and atherosclerosis. There is also some evidence for CMV involvement in cancer. Viral genomes and proteins have been found in tumor cells and may enhance tumor development and growth. On the beneficial side, individuals who are CMV seropositive have been shown to have enhanced T-cell repertoires and improved responses to influenza vaccination.
TREATMENT
Despite considerable progress, additional safe and effective therapies are needed for all affected populations (pediatric patients, pregnant women, the immunocompromised). The development of ganciclovir, foscarnet, cidofovir, and valganciclovir demonstrated that beneficial direct-acting antiviral (DAA) therapy was possible in patients with human immunodeficiency virus infection, in transplant recipients, and even in some congenitally infected infants. However, the significant associated toxic effects limit the drugs’ utility for all but the sickest patients.
The search for other DAAs remains intense, to overcome the toxic effects and the emergence of resistance to the marketed antivirals. Letermovir, a CMV terminase inhibitor, has been approved for prophylactic use in hematopoietic stem cell transplant (HSCT) recipients and is the first new CMV therapeutic approved since 2001. Maribavir, a UL97 kinase inhibitor, is in phase III trials in HSCT recipients and in solid organ and HSCT recipients with CMV infections that are refractory or resistant to ganciclovir, foscarnet, valganciclovir, or cidofovir. Brincidofovir, a polymerase inhibitor and lipid-ester prodrug of cidofovir, showed promising efficacy in a phase II trial in HSCT recipients; however, a phase III trial did not meet its primary end point. Filociclovir, which has activity against both the viral polymerase and UL97 kinase, is in early clinical trials.
For transplant recipients, the strategies of classic prophylaxis therapy versus preemptive therapy are under investigation. Preemptive therapy has the advantages of treating those at highest risk and limiting exposure to the toxic effects associated with currently available therapies. Preemptive therapy may also have the advantage of allowing a small amount of virus replication, thus stimulating the immune system to better control CMV disease. A possible downside of preemptive therapy is that it may not adequately control compartmentalized disease that is not adequately correlated with viremia and may not effectively prevent indirect effects of CMV infection.
Immunoglobulins are another class of DAAs under investigation for clinical utility. Currently, Cytogam, a high-titered anti-CMV pooled human immunoglobulin, is used to prevent CMV infection in solid organ transplant recipients. A few early-phase clinical studies have raised the intriguing possibility that an appropriate immunoglobulin prophylactic therapy for seronegative pregnant women might prevent or ameliorate congenital CMV (cCMV) infection in their fetuses. The clinical potential of immunoglobulin therapy, as monotherapies or cocktails, has stimulated research into monoclonal antibodies. Not surprisingly, target selection is key. As with other DAAs, the resistance can develop. Some viral antigens have strain variations and consequently are not suited for a broad-CMV-spectrum agent. Moreover, some candidate monoclonal antibodies have been shown to inhibit infection of epithelial cells but not infection of fibroblasts.
There is also intense research into the potential of 2 other categories of therapies: host-targeted antivirals and adoptive CMV-specific T-cell therapy. Host-targeted antivirals target a host function that the virus subverts for its own replication or pathogenesis. These generally lack the potency of DAA but have several potential advantages, including lack of susceptibility to the development of resistance and possible activity against a broad spectrum of viruses. The sirtuins, which are involved in the regulation of cellular metabolism, are a promising target for this class of antivirals. Adoptive CMV-specific T-cell therapy has demonstrated positive results in transplant recipients with drug-resistant or refractory CMV disease, but it is currently limited by limited donor availability and the time required to generate the cells. New, “off-the-shelf” virus-specific T-cell banks are a promising development and should permit future clinical trials.
PREVENTION
Currently, a number of vaccines are in clinical and late preclinical development, with different strategies and differing planned clinical paths. Most involve the use of a mix of viral antigens such as glycoprotein B (gB) to target entry into fibroblasts, pp65 to elicit CD4+ and CD8+ T-cell immunity, and the pentameric complex to target entry into epithelial/endothelial cells and elicit highly potent neutralizing antibodies. Because of the varied manifestations of CMV disease in different target populations, there is the real possibility that a single vaccine might not provide the optimal level of protection for fetuses, transplant recipients, and patients with other forms of immunocompromise.
Many features of the natural history of CMV complicate vaccine design. The first is the lack of an established correlate of immunity. However, other vaccines have been successfully developed without a clear understanding of how they worked. Furthermore, the ability of CMV to superinfect a seropositive individual suggests that sterilizing immunity is probably not possible. At the same time, complete immunity may not be needed, because data from studies of both congenital infection and transplantation suggest that preexisting immunity can prevent or ameliorate the severity of disease. The ability of the virus to spread cell to cell is presumably important to fetal infection and probably requires that an effective vaccine be able to elicit a cellular response.
DIAGNOSIS
As with treatments and vaccines, diagnosis of CMV infection varies with the patient population. For solid organ transplant or HSCT recipients, diagnosis is primarily based on CMV quantitative polymerase chain reaction of blood or plasma. Although CMV quantitative polymerase chain reaction has improved the diagnosis of CMV infection, the variability among assays has led to the recent establishment of international standards. Evaluations of methods for assessing CMV DNA that do not require a clinic or laboratory visit (eg, self-collected dried blood spots) represent a next phase in follow up of patients after transplantation, particularly to assess late CMV disease. For pregnant women, diagnosis of maternal infection or reactivation is particularly challenging. To date, most testing has been performed to identify the serostatus of the pregnant woman, but the serostatus testing is typically only useful if the serologic status is known before pregnancy. Moreover, the current serologic end points (seroconversion, positive immunoglobulin M, and low avidity index) have limitations; therefore, routine antenatal serologic screening is not recommended to diagnose human CMV status. Large prospective studies may be needed to develop and validate algorithms for CMV testing in pregnancy.
Laboratory diagnosis of cCMV infection in newborns includes virus isolation from urine or saliva (long the reference standard), or more recently, quantitative nucleic acid testing of saliva and/or urine. However, except for the most severely affected, most infants with cCMV infection are not identified at birth. Universal screening of newborns has been recently adopted by several states in the United States. Testing of dried blood spots, which are collected from all newborns in the United States, may have some utility in universal screening, but improved and standardized methods for DNA extraction and testing are needed. Detection of viral DNA in urine or saliva is currently the clinical standard; however, the timing of sample collection from the infant is critical, because CMV can be present in breast milk. Therefore, breastfeeding may yield a false positive saliva sample.
Continued development of diagnostic methods to detect CMV infection in all of these patient population is essential. Beyond CMV DNA detection, clinically applicable assays of cell-mediated immune responses and new imaging technologies for prognostic assessments are also needed.
RESEARCH GAPS
Despite the extensive characterization of the composition and expression of the large CMV genome, much research remains to be done to identify the full complement of viral open reading frames and their role in pathogenesis and infection. Although cellular receptors have been identified, they are specific to separate cell types. More effort is needed to determine both their impact on pathogenesis and the possibility that the full complement of receptors and accessory proteins remains to be identified. The genomic intrahost diversity of CMV genomes is extensive and reminiscent of that seen with several RNA viruses. Is this important to pathogenesis and strategies for control of infection and disease? Importantly, much remains to be discovered on the interaction of CMV-expressed functions and cellular functions. Interactions with host inflammatory and immune systems clearly affect the outcomes of infection.
For seropositive women, information on the role of reinfection versus reactivation in the development of cCMV would certainly affect strategies for prevention. Among infants born with asymptomatic cCMV, sensorineural hearing loss will develop in approximately 15%. The identification of prognostic markers is key to avoiding unnecessary treatment in most of these infants. Prognostic markers more sensitive than viremia are also needed for optimal management in transplant recipients. For CMV involvement in other diseases, it is important to define both strain variation and the definition of viral gene expression during latency. Because CMV disease may occur in several organ systems, the compartmentalization of the effects of infection and the variety of responses in different compartments in a single host are important to understand.
Although several new drugs are in clinical and preclinical development, the research pipeline is too thin. Considering the fragility of many of the target populations and the complexity of the virus-host interactions, the search for therapeutics with an acceptable safety profile that can block viral replication and/or mechanisms of pathogenesis is extraordinarily challenging. If a panherpes drug could be identified, that would probably be more attractive to the pharmaceutical industry. In addition, because achieving adequate volunteers for new drug evaluation can be challenging for many patient populations, it is critical that preclinical and clinical studies and trials are designed carefully and have appropriate end points so that the results accurately answer the clinical question. The identification of reliable biomarkers is important for this activity. As biologic therapies that modify the host immune system are becoming more widely used, it will be necessary to monitor recipients for the unintended consequence of problems with CMV reactivation or infection.
The major knowledge gap for vaccine development is the definition of correlates of immunity and which immune parameters are most important for protection from new infection versus reactivation. Exploration of the actual and potential roles of passive immunity are important for all CMV disease manifestations. The optimal vaccine strategy, unfortunately, may differ among patient populations, which may result in either a “compromise” vaccine or the need to develop 2 or more separate vaccines. The duration of immunity is a significant issue, especially for vaccines that target cCMV.
For the diagnosis of CMV, important research areas include evaluations of the utility of imaging technologies for assess localized disease or disease severity, the development and validation of assays for cellular immune response and improved diagnostic CMV assays and algorithms. Finally, large-scale clinical and epidemiologic studies will be important for assessing the impact of CMV infection or reactivation on cCMV disease and will form the basis for a vaccine implementation strategy.
Note
Supplement sponsorship. This supplement was sponsored by NIAID and NICHD.
Potential conflicts of interest. All authors: No reported conflicts. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.