Outside of improvements in immunosuppressive therapy, the control of cytomegalovirus (CMV) infection in solid organ transplantation has been considered as the most significant advance in organ transplantation to have taken place in the past 20 years [1]. As a result of prevention of CMV infection, both the direct and the indirect effects secondary to CMV replication have led to improved outcomes. CMV disease and disseminated CMV disease have virtually disappeared as clinical occurrences in most organ transplant recipients, and the effects that are indirectly associated with CMV disease, namely rejection and opportunistic infection, have been reduced dramatically [2–5]. As a consequence, there is improved survival and decreased cost for patients undergoing solid organ transplantation [6–9]. Out of recognition that control or prevention of CMV disease is important for graft and patient survival, 2 strategies have emerged, both of which are outlined in the article by Singh [10]—namely, the use of prophylaxis and the use of preemptive therapy. The terminology has been a bit confusing to those outside the field, because 2 different preemptive strategies have been used. One involves suppression of viral replication after detection of CMV load [11–13]. The other involves administration of “prophylaxis” to a subset of individuals who are at very high risk for reactivation, because of an increased likelihood for CMV disease associated with increased levels of immunosuppression [14]. This latter strategy was first termed “preemptive therapy” by Robert Rubin ∼15 years ago [15], but it might be more properly termed “select prophylactic therapy,” because prophylaxis is initiated before the documentation of viral replication [16]. Thus, we have 2 dominant strategies, as well as a minor modification of the prophylactic strategy, that have been the primary modes of preventing CMV infection and disease. The thesis put forth by Singh [10] is that patients receiving CMV prophylaxis may be at increased risk of developing late-onset CMV disease, which may be serious and even fatal. Furthermore, Singh [10] asserts that the use of prophylaxis may be associated with CMV resistance to ganciclovir or valganciclovir and that the failure to allow the patient to develop a cellular immune response in relation to viral replication permits the development of late onset or relapsing disease. The most basic tenant of and the theory behind preemptive treatment is that a small rate of CMV replication is a good thing. The ideal prophylactic agent should be one that is highly effective, nontoxic, easily administered, and inexpensive. Unfortunately, valganciclovir or ganciclovir—although highly effective for preventing CMV replication—are not ideal agents. These drugs are very expensive and potentially toxic. Furthermore, as Singh points out, prophylaxis is associated with the development of drug-resistant CMV and with late-onset CMV disease after discontinuation. There are also very limited data involving solid organ transplant recipients to support the concept that there is abrogation of the immune response to CMV replication [17]. Although all of these issues have some merit, there would not be any debate about the appropriate strategy if we had an agent that was as effective, easily administered, and inexpensive as trimethoprim-sulfamethoxazole is for preventing reactivation of Pneumocystis jiroveci infection in immunosuppressed populations. The use of trimethoprim-sulfamethoxazole has been generally adopted in all transplantation centers because it has all of the attributes of an effective prophylactic regimen [18]. Let us examine the evidence concerning the development of ganciclovir resistance in CMV following the use of ganciclovir prophylaxis. To date, emergence of drug-resistant CMV has been a very rare phenomenon in recipients of solid organ transplants [19, 20], and, for the most part, it has been reported in case series or case reports [21]. This has been the situation for the past 6 years, even with the use of oral ganciclovir, which has significantly less bioavailability than valganciclovir [22]. There are several characteristics associated with the emergence of drug-resistant CMV that are clearly understood. First, the groups at highest risk for infection seem to be recipients of kidney-pancreas or lung transplants [20, 22]. Such risk might be understandable for the latter group, because the CMV load in the lung is thought to be higher than that for any other organ [23]. Why kidney pancreas transplant recipients are at such high risk is not clear, but the reason may be associated with the pharmacokinetics of ganciclovir therapy in this population. Second, most patients in whom ganciclovir-resistant CMV emerges are seromismatched (i.e., CMV-seronegative recipients of transplants from CMV-seropositive donors) [20, 21]. Third, it seems that patients who receive multiple courses of ganciclovir for prophylaxis or therapy and then relapse after treatment are at particularly high risk [20, 21]. There is evidence that use of a preemptive strategy for the high-risk subgroup of lung transplant recipients is associated with the emergence of antiviral-resistant CMV [24]. In a cohort of lung transplant recipients who were at high-risk (i.e., seromismatched recipients) or intermediate risk (i.e., CMV-seropositive recipients) for CMV infection and disease, the risk of emergence of drug-resistant CMV among those who received preemptive therapy was ∼10% [24]. In a trial of solid organ transplant recipients limited to patients who were CMV seromismatched that compared valganciclovir prophylaxis with ganciclovir prophylaxis, no patient in the valganciclovir arm had CMV with the resistance mutation UL 97, compared with 1.9% in the ganciclovir arm [25]. Furthermore, among patients in this study who received oral ganciclovir prophylaxis and developed CMV disease, 6.1% had CMV with the UL 97 mutation. One operating assumption is that use of a drug with greater bioavailability will reduce the likelihood of the emergence of the UL 97 mutation [25]. However, it should be noted that there was evidence of CMV replication in ∼40% of patients 6 months after discontinuation of valganciclovir prophylaxis, which underscores the need to either monitor such patients for symptoms or for evidence of viral replication or the need for more-prolonged prophylaxis [26]. Understanding why some patients develop viral replication and others do not would be important to formulate the most effective strategy. There is no question that there has been change in time of onset and changing epidemiology of late-onset CMV disease [20, 27]. I and others have also noted that late-onset disease may have very atypical features [27–29]. Some of the changes in the epidemiology of CMV disease may be due to the use of prophylaxis, but some may be due to changes in the type of immunosuppressive therapy or in our ability to recognize late-onset CMV disease. In fact, I and others have observed that disease associated with a very late onset (often even years after transplantation) is occurring at a greater incidence than was previously appreciated [29]. The development of the immune response in solid organ transplant recipients after a small amount of viral replication is of interest but is not well understood. It is clear that a comparison of solid organ transplant recipients and bone marrow transplant recipients with respect to late-onset disease or failure of host immune response is not appropriate. These are vastly different biological models for several reasons. First, donor immunity is helpful in the case of bone marrow transplantation; thus, a CMV-seropositive bone marrow donor transplant may actually reduce the level of CMV replication in and even transfer humoral immunity to the recipient [30, 31]. Second, the immune response itself and the immunosuppressive regimens used in these populations are very different. There are a number of risk factors other than use of prophylaxis that are associated with the development of late-onset CMV disease in solid organ transplant recipients, including the need to administer treatment for rejection, retransplantation, and the development of hypogammaglobulinemia [32, 33]. Furthermore, in solid organ transplant recipients, development of CMV disease may be associated with specific HLA-DR haplotypes [34, 35]. There is even evidence that late-onset CMV disease may occur in association with preemptive therapy [36] and that late-onset CMV infection may occur despite CMV-specific CD8+ T cell activation [37]. The 2 strategies that have emerged—preemptive therapy and prophylaxis—have only been compared head-to-head in small trials designed primarily to show their equivalence in preventing CMV disease. In 2 comparative trials [38, 39], a number of deficiencies can be found. First, they were not powered to show a difference in the frequency of death, graft rejection, or opportunistic infection; the primary end points were CMV disease onset and treatment cost. Second, in one of the trials, an inferior agent for antiviral prophylaxis (acyclovir) was compared with a superior agent for anti-CMV prophylaxis (ganciclovir) [38]. One conclusion that can be drawn is that the preemptive strategy is less costly because fewer courses of antiviral therapy are administered and the cost of monitoring is not as expensive as the cost of drug. In addition, the preemptive strategy exposes fewer patients to the potentially toxic effects of the drug. However, in the subset at highest risk for diseases (the CMV-seromismatched group), most patients will need some intervention [36], and it has been argued that prophylaxis is the most cost-effective strategy [40]. Furthermore, there is also evidence that CMV resistance may occur in association with a preemptive antiviral regimen [24]. Intuitively, one might think that suppression of viral replication with prophylaxis might reduce the likelihood of indirect effects of CMV disease, such as graft rejection and opportunistic infection. As expected, a recently presented meta-analysis of antiviral prevention of CMV disease in solid organ transplant recipients demonstrates just such a benefit [41]. In contrast to the preemptive strategy, the prophylactic strategy showed a significant reduction in graft loss, mortality, and the development of bacterial or fungal opportunistic infection. The preemptive strategy only demonstrated a significant decrease in graft rejection. Both strategies demonstrated a reduction in CMV disease. For patients in the CMV-seromismatched group, the prophylactic strategy was the only strategy to demonstrate a reduction in CMV end-organ disease. Therefore, when one examines the impact of CMV prophylaxis on the many indirect effects of CMV disease and also assesses the impact on patients who have a high serological risk of disease, the evidence suggest that the prophylactic strategy is significantly superior to the preemptive strategy [41]. This is consistent with the hypothesis that the indirect effects of CMV infection are exceedingly important components in any strategy and that viral suppression rather than permissive viral replication followed by subsequent suppression will improve the outcomes for solid organ transplant recipients. One can acknowledge that the current strategy of prophylaxis-based CMV prevention is imperfect, but the ease of administration and the general overall effectiveness of this approach are well accepted and have been documented in many trials over the past decade. The superior pharmacodynamic profile of valganciclovir is the most attractive reason to use this drug as prophylaxis [22, 26]. However, clinicians need to be aware of late-onset CMV disease and its atypical manifestations [27, 29]. The occurrence of atypical or late-onset CMV disease in populations that do not have the highest serological risk is extremely rare but can occur. It would be very difficult to develop a trial with the power to prove that preemptive therapy is equivalent to prophylaxis for preventing the indirect effects of CMV infection (including survival and rejection), because of the impressive outcomes achieved in most solid organ transplant programs [7, 42, 43] and because it is probable that a sufficient sample size cannot be achieved, even with a multicenter trial. Although the debate may continue, the evidence suggests that, when both the indirect as well as the direct effects of CMV infection are considered, prophylaxis is superior to preemptive therapy, and the data also suggest that prophylaxis is the most cost-effective strategy [44]. Potential conflicts of interest. D.R.S. has recently received research funding from Roche, MedImmune, Merck, Bayer, and AstraZeneca; has served as a consultant for MedImmune; and has served on speakers' bureaus for Roche, Bayer, Merck, Cubist.
No takes yet. Share an insight, caveat, or question.
David R. Snydman (2005) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: