CMV-RNAemia as new marker of active viral replication in transplant recipients

,

prophylaxis, CMV-RNAemia resulted positive in 6/25 (24%) episodes (Table 1).Among these, 4/6 (66.7%) were related to clinically significant CMV infections (based on both clinical and laboratory findings), treated by pre-emptive therapy.In 2/4 (50%) cases, genotypic LMV resistance (LMV-R) was documented.Considering the group receiving LMV off-label, six out of seven episodes (85.7%) showed positive CMV-RNAemia (Table 1), and in 3/6 cases (50%), LMV-R was found, leading to switching the treatment to other anti-CMV agents: maribavir, foscarnet, or immunoglobulins.In the remaining three cases, the LMV treatment was continued and combined with a reduced immunosuppres sive therapy, until complete viral clearance (two consecutive CMV-DNAemia negative results).In the 12 out of 32 (37.5%) episodes in which CMV-RNAemia was detected during LMV administration, the active viral replication was documented by CMV-vire mia and/or DNase tests.These additional methods also confirmed the CMV-RNAemia negative results in the remaining 20/32 (62.5%) episodes, suggesting that the positive b LMV-off-label due to ganciclovir resistance (three patients receiving solid organ transplant), severe neutropenia (two pediatric hematopoietic stem cell transplant recipients), and positive CMV-DNAemia results when starting LMV (two adult hematopoietic stem cell transplant recipients).c The patients receiving therapy with (val)ganciclovir (n = 13) or foscarnet (n = 1).d CMV-DNAemia-positive episodes: at least two sequential positive CMV-DNAemia results in whole blood or one more than 300 copies/mL; a mean of 5.4 samples/episodes [±3.5 standard deviation (SD)] was analyzed.
e Conversion factor from copies/mL to international units/mL is 0.46.f CMV-RNAemia-positive episodes: episodes with at least one positive result.
g Spearman correlation, CMV-DNAemia and CMV RNAemia levels under the lower limit of quantification (300 and 30 copies/ml) were considered equal to 150 and 15 copies/ml, respectively.h Pearson correlation, CMV-DNAemia and CMV RNAemia levels under the lower limit of quantification (300 and 30 copies/ml) were considered equal to 150 and 15 copies/ml, respectively.
i The specificity of 100% was also found in 100 additional plasma samples from transplant patients for whom CMV reactivation was ruled out by negative CMV-DNAemia results (data not shown).
CMV-DNAemia in these cases was possibly due to abortive infections.Finally, CMV-RNAe mia was positive in all 15 episodes from 14 patients receiving pre-emptive therapy (Table 1).Analyzing all the 227 positive samples (Table 1), the CMV-DNAemia levels were higher in specimens positive for CMV-RNAemia than in those negative (P < 0.001, Fig. 1).Considering the three groups of patients, lower median CMV-DNAemia values were found in cases receiving LMV, as prophylaxis or off-label, than in the pre-emptive group (P < 0.001, Mann-Whitney test).The results confirmed that detectable low DNAemia levels during LMV administration may reflect abortive rather than productive infection.These data also suggest that CMV-RNAemia could be useful for detecting active CMV replication in patients receiving antiviral therapy, especially with drugs that do not act on the viral DNA polymerase, such as LMV.
Considering the 27 episodes positive for both markers (Table 1), the CMV-DNAemia and CMV-RNAemia peaks were reached simultaneously, with median levels equal to 11,754 copies/mL (IQR: 7,386.5-26,174)and 81 copies/mL (IQR: positive <30-489.5),respectively.Interestingly, during the descending phase, negative results were obtained earlier with CMV-RNAemia than with CMV-DNAemia (mean time 7 days before, ±6.9 standard deviation), proving more rapidly an efficient viral clearance, as shown in a representative case detailed in Fig. 2. As reported in literature, during this phase, CMV-DNAemia could be detected due to the presence of free viral genome fragments that are released after infected cell degradation and are not necessarily associated with infectious viral particles (3,4).These observations, together with the previous consider ations, show that CMV-DNAemia alone may not be an accurate marker of active CMV replication, especially in patients undergoing LMV.
In this study, CMV-DNAemia was measured in WB samples since previous investiga tions suggested this blood compartment as a preferable clinical sample for monitoring CMV infection post-transplantation (3,4,8).As RNA is an unstable and fragile molecule, plasma samples were immediately frozen at −80°C and analyzed within 3 weeks.At prospective and retrospective testing of CMV-RNAemia in a small group of samples, no significant differences were observed (data not shown), confirming the absence of degradation during storage.
In conclusion, CMV-RNAemia, detected by a standardized user-friendly assay and an automated and integrated system, together with CMV-DNAemia, could provide accurate information on viral load kinetics during post-transplant monitoring of CMV infection, especially in patients receiving LMV.Further studies with larger numbers of samples, including patients undergoing therapy with new anti-CMV drugs such as maribavir, are needed to confirm these preliminary data.

TABLE 1
Characteristics of study population, infective episodes, and samples analyzed