Resistant/Refractory Cytomegalovirus Retinitis in Solid Organ and Hematopoietic Stem Cell Transplant Recipients: The Cause‐or‐Consequence Dilemma

Dear editor, Cytomegalovirus (CMV) retinitis (CMVR) is a severe and potentially blinding end-organ disease in the immunocompromised host. CMVR is a well-known complication in patients with acquired immunodeficiency syndrome (AIDS) [1, 2]. However, CMVR incidence and risk factors in non-HIV immunosuppressed individuals, particularly in solid organ transplantation (SOT) and hematopoietic stem cell transplantation (HSCT) recipients, remain unclear. Its prevalence in SOT/HSCT patients is limited and derives mainly from retrospective studies, ranging between 7%–15% [3, 4]. However, CMVR diagnosis is challenging given that up to 50%–66% of patients can be asymptomatic at diagnosis [5]. In addition, there are no indications or recommendations from guidelines regarding ocular involvement screening in SOT/HSCT populations with concurrent CMV infection or disease [6, 7]. Retinal localization in refractory-resistant CMV (RR-CMV) infection should be seen as an even more challenging scenario due to the limited therapeutic options and the poor ocular penetration of the available antiviral drugs [6, 8, 9]. We present three clinical cases of RR-CMV retinitis in SOT/HSCT recipients (Table 1). CMV retinitis diagnosis from transplant (in months) D+/R+ prophylaxis with (V)GCV 25 days 53.199 cp/mL 7 Antiviral resistance test: -December 2016: UL97 (M460V) -January 2017: UL97 (M460V) and UL54 (Q578H + E756 D) Leflunomide, CMV-Ig Clinical: loss of vision. Virological: CMV-DNA on ocular samples and blood was negative in 2 months D−/R+ pre-emptive strategy 72 days 36.651 cp/mL Clinical: complete resolution in 3 months. Virological: CMV-DNA on ocular samples and blood was negative in 1 months D−/R− prophylaxis with GCV 47 months 16.751 cp/mL FOS, MBV from October to December 2024(3 months) Stop mycophenolate CMV-Ig Clinical: loss of vision left eye. Virological: CMV-DNA on blood was negative in 1 months A 40-year-old female underwent multivisceral transplantation (liver/pancreas/bowel) in 2016 for bowel insufficiency. CMV serostatus was D+/R+. Ganciclovir (GCV) prophylaxis was performed. After 1 month from SOT, blood CMV-DNA levels rose (53.199 copies/mL) and GCV was started. CMV drug resistance test was conducted due to high viral loads despite GCV appropriate treatment. Mutation on UL97 gene (M460V) was detected. Foscarnet (FOS) was then started. After 22 days of FOS therapy, poor viral clearance was observed. Therefore, a new CMV drug resistance test was performed, revealing both mutations on UL97 and UL54 (Q578H and E756D) genes, conferring resistance to GCV, FOS, and cidofovir (CDV). Lately, leflunomide and CMV-Immunoglobulins (CMV-Ig) were administered with resolution of infection. At 7 months from SOT, rapid onset of visual loss was reported. Bilateral CMVR with retinal hemorrhage, posterior chamber vasculitis, and optic nerve atrophy were confirmed at ophthalmological examination and fluorescein angiography. CMV-DNA elevation in blood was observed reaching to 10.199 copies/mL. Therefore, a new course of combination treatment with leflunomide and CMV-Ig was started, achieving blood CMV DNA clearance within 2 months, while bilateral optic nerve atrophy worsened, leading to complete blindness. A 50-year-old female underwent allo-HSCT in 2013 for T-leukemia/lymphoblastic lymphoma. CMV serostatus was D−/R+. Pre-emptive CMV strategy was adopted. GCV therapy was started 72 days post-HSCT for blood CMV-DNA elevation (36.651 copies/mL). Due to persistent high CMV-DNAemia despite appropriate antiviral therapy, resistance test was performed and no mutations were revealed; therefore, the infection was defined as refractory CMV infection. Rapid onset of visual loss occurred, and ophthalmological examination revealed early retinal ischemic damage. FOS was, therefore, started together with intravitreal corticosteroid injections, leading to viral clearance and complete resolution of retinitis signs within 3 months. A 62-year-old male underwent a kidney transplant in 2020 for chronic kidney disease due to glomerulosclerosis. CMV serostatus was D−/R−. Due to relapsing episodes of CMV infection/disease, a very long GCV/VGCV-exposure (47 months) was observed. Despite antiviral therapy, in 2024, unilateral acute retinal necrosis occurred with evidence of intravitreal CMV localization (25,000.000 copies/mL). Resistance test detected L595F mutation and 596–600 deletion on UL97 gene, conferring resistance to GCV/VGCV. FOS systemic therapy was started but later shifted to maribavir (MBV) for drug intolerance. Concomitantly, mycophenolate was discontinued. Due to poor MBV penetration in the central nervous system (CNS), intravitreal FOS administrations and CMV-Ig were also administered. One month later, virological clearance on blood was achieved, and ophthalmologic controls excluded progression of CMVR. We report three cases of retinitis due to RR-CMV in SOT and HSCT recipients. Current evidence suggests that CMV retinitis is a relatively uncommon complication in these populations [3]. However, the rate of asymptomatic or poorly symptomatic CMVR, together with the lack of clear recommendations for ocular screening in transplant recipients, suggests that its prevalence may be underestimated [5-7]. Furthermore, there are no data establishing a correlation between the development of retinitis due to RR-CMV strains in SOT or HSCT recipients. Conversely, GCV-resistant CMVR have been reported among patients with AIDS [2, 10, 11]. In these reports, prolonged exposure to GCV, high viral loads, and variable drug levels (resulting from poor absorption, underdosing, or adherence issues) were identified as risk factors for the development of resistance. Management of CMVR is challenging, and the emergence of CMV genetic mutations conferring antiviral resistance to first-line antiviral treatment (e.g., GCV) further complicates its treatment due to the limited therapeutic options available for RR-CMV disease [9]. In particular, the efficacy of MBV in treating CMV retinitis appears to be limited because of its poor penetration into ocular tissues [6, 12]. Consequently, FOS, administered either intravenously or through local injections, often remains the only viable therapeutic option despite its well-recognized toxicity profile and probably worse outcome compared to GCV [13]. All this considered, a “cause-or-consequence” question arises: does refractory CMV infection predispose to ocular involvement, or does localization of infection within a sanctuary site such as the eye promote the emergence of antiviral resistance? One possible explanation is that CMV retinitis may result from viral escape during treatment with second-line antiviral agents that have limited ocular penetration in the setting of systemic infection [14]. Conversely, infection confined to a sanctuary site like the eye may, in turn, facilitate the development of resistance to first-line antiviral drugs. It is difficult to answer this question for several reasons. First, there are currently no validated recommendations for routine screening of ocular complications in SOT or HSCT recipients with CMV infection. Such recommendations might be particularly valuable in patients with RR-CMV infection, in whom early ophthalmologic evaluation could enable prompt diagnosis and timely initiation of therapy, potentially preventing the worrisome and irreversible consequences of CMV retinitis [2, 15]. Furthermore, data on the clinical manifestations of RR-CMV infection remain limited, being mostly derived from small retrospective cohorts without a specific focus on ocular involvement [4]. Consequently, the risk factors for RR-CMV retinitis in the SOT and HSCT population remain unclear, underscoring the need for further prospective studies. In conclusion, given the uncertain incidence of CMV retinitis, its potentially severe ocular sequelae, and the limited therapeutic options available, implementing routine screening for retinal involvement may be warranted in SOT and HSCT recipients with RR-CMV infection. Early diagnosis and timely initiation of appropriate therapy may lead to significant improvements in ophthalmologic outcomes in this population. The authors have nothing to report. The authors declare no conflicts of interest.