Caution Is Required before Recommending Wide Use of Letermovir as Salvage Therapy for Cytomegalovirus Diseases

LETTER

Turner et al. recently reported the use of letermovir in 4 solid-organ transplant recipients with ganciclovir-resistant cytomegalovirus (CMV) and retinitis (1). Because all 4 patients showed clinical and fundoscopic improvement with resolution of retinitis, the authors concluded that this case series highlights promising features of letermovir as salvage therapy for refractory CMV infection.

Although this conclusion may be true, healing of retinitis could be due, at least in part, to the intravitreal foscarnet and/or ganciclovir injections given in 3 of the 4 cases. Indeed, previous studies suggested that successful local treatment of CMV retinitis (with ganciclovir implants or intravitreal injections and/or foscarnet intravitreal injections) may be achieved in the absence of concomitant systemic antiviral therapy (2–4). Another explanation for the contrasting result between resolution of retinitis and persistence of viremia is that the CMV species involved in the retinitis may have remained susceptible to letermovir. The authors did not report the genotypic characteristics of the CMV present in aqueous fluids. However, previous comparisons of CMV genotypic resistance profiles in concomitant samples obtained from blood and aqueous or cerebrospinal fluids showed that viral compartmentalization may occur during CMV diseases, with isolation of wild-type viruses in aqueous and/or cerebrospinal fluids samples and resistant CMV in blood (2, 3, 5, 6).

Turner et al. report genotypically documented resistance to letermovir (mutation C325F in the UL56 gene) in 2 patients among the 3 who failed to control CMV viremia during letermovir treatment. We agree with their conclusions that this report emphasizes the important concern for emergence of resistance while receiving letermovir. However, the authors suggest that the observed cases of resistance may have resulted from selection of resistant subpopulations of CMV rather than as a consequence of a low barrier to resistance. Because all the patients were letermovir naive at the time of diagnosis of their retinitis, we believe that this hypothesis seems unlikely. Indeed, previous in vitro studies suggested that naturally occurring polymorphisms associated with reduced letermovir sensitivity are very unlikely in cytomegalovirus field isolates (7–9). Therefore, we suggest that the selection of the C325F mutation described in 2 patients of this case series probably did not predate the letermovir treatment and has been selected during virological failure, as previously described in other cases of letermovir use for salvage treatment in transplant recipients with CMV disease (10–11).