Use of Letermovir for Salvage Therapy for Resistant Cytomegalovirus in a Pediatric Hematopoietic Stem Cell Transplant Recipient

Abstract

We present here the first published use of letermovir for the treatment of resistant cytomegalovirus (CMV) in a pediatric patient. A 14-year-old girl underwent a double unrelated umbilical cord blood transplantation to treat her sickle cell disease (hemoglobin SS) and developed ganciclovir-resistant CMV DNAemia with end-organ involvement that was treated successfully with a combination of foscarnet and letermovir. After she was transitioned to letermovir monotherapy for secondary prophylaxis, she developed recurrent DNAemia with laboratory-confirmed ganciclovir, foscarnet, and letermovir resistance.

Pediatric data regarding the use of letermovir for prophylaxis and treatment in pediatric patients are lacking. If letermovir is used for salvage treatment purposes, combination therapy should be considered for patients at high risk.

Cytomegalovirus (CMV) infection as a cause of serious complications in hematopoietic stem cell transplant (HSCT) recipients, including graft failure, has been well reported [1, 2]. Sustained DNAemia can also lead to the development of resistance to antiviral drugs [3]. Novel therapies such as letermovir have broadened the treatment options for resistant CMV disease. To our knowledge, this is the first published case of letermovir use for treating resistant CMV disease in a pediatric patient.

CASE PRESENTATION

A 14-year-old girl with a past medical history of sickle cell disease (hemoglobin SS) and progressive neurologic sequelae underwent a planned double unrelated umbilical cord blood transplant after myeloablative preconditioning with fludarabine, busulfan, and cyclophosphamide. The cord blood cells tested negative for CMV infection. The recipient was known to be CMV seropositive before her transplant. After the transplant, graft-versus-host disease (GvHD) prophylaxis consisted of cyclosporine and mycophenolate. The patient showed early signs of engraftment on day 7 but did not fully engraft until day 24. In the periengraftment period, she developed fevers and a nonspecific skin rash. An infectious disease workup during this time revealed human herpesvirus 6 (HHV6) and CMV DNAemias, which were treated with foscarnet monotherapy. The remainder of her infectious disease workup results were negative. Her skin rash was consistent with cutaneous GvHD; she was treated with systemic corticosteroids, and her skin findings improved moderately and her fevers resolved within 1 week of the engraftment. A comprehensive list of her serum CMV trends, antiviral therapy, GvHD classification, and concurrent immunosuppression are detailed in Figure 1.

Figure 1.

Cytomegalovirus (CMV) viremia versus time with antimicrobial immunosuppressive therapies. Elapsed time is in reference to the patient’s transplant date (day 0). Engraftment occurred on day 24. CMV resistance testing was performed on days 48, 90, and 257 after transplantation and is denoted by vertical arrows. Note the worsening CMV viremia despite aggressive antiviral therapy within the setting of significant graft-versus-host disease (GvHD) that necessitated multifaceted immunosuppression, including the use of alemtuzumab. Abbreviations: IV, intravenous; PO, oral; q8h, q12h, q24h, every 8, 12, and 24 hours, respectively.

Through day 30, the patient remained clinically stable despite persistently low-level HHV6 and CMV viral replication. Foscarnet was reduced from treatment to maintenance dosing, and serum viral polymerase chain reaction monitoring was continued. On foscarnet maintenance dosing, the patient had a consistently rising CMV DNAemia level by day 45, but her HHV6 levels remained either undetectable or detectable at <188 DNA copies per mL. Foscarnet maintenance therapy was stopped, and intravenous (IV) ganciclovir therapy (5 mg/kg every 12 hours) was initiated. Her CMV DNAemia level continued to rise despite the use of treatment-dosed ganciclovir. Initial CMV resistance testing results on day 48 were negative. The IV ganciclovir dosage was increased from 5 to 7.5 mg/kg every 12 hours to optimize therapy. On this ganciclovir dose, her CMV DNAemia level reached a nadir of 707 IU/mL (log 2.8 IU/mL) by day 75.

Despite achieving this viremic nadir, we found evidence in the patient of worsening GvHD refractory to an 8-week course of methylprednisolone (1–2 mg/kg per day). Her immunosuppressive regimen was modified, including the substitution of tacrolimus for cyclosporine and addition of mesenchymal stromal cells and infliximab. On day 90, the patient was having daily recurrent fevers and progressively increasing stool output. The results of a broad infectious disease workup, including stool studies, were negative other than persistent CMV DNAemia. Repeat resistance testing on day 90 revealed an A594V UL97 mutation conferring resistance to ganciclovir. However, resistance mutations against foscarnet were not detected. She was transitioned back to IV foscarnet (60 mg/kg every 8 hours). A colonoscopy with biopsies was performed. Colonic biopsy results were positive for increased epithelial cell apoptosis, crypt injury with focal crypt loss, and rare cells suspicious for viral cytopathic effect with positive CMV immunohistochemical staining. The ultimate pathological findings were consistent with CMV enterocolitis and concurrent grade IV GvHD. Alemtuzumab (Campath [Genzyme, Cambridge, MA, USA]) was added on day 95 to treat her steroid-refractory acute GVHD.

The patient continued to have intermittent fevers, copious loose stools, and moderate transaminitis after the addition of alemtuzumab and despite changes in her antiviral therapy. Her CMV DNAemia peaked at 816 281 IU/mL (log 5.9 IU/mL) on day 99. As a result of her worsening DNAemia despite the use of foscarnet, IV letermovir (480 mg once daily) was added for dual antiviral salvage therapy. The results of another workup, including a dilated eye examination, were negative for evidence of other CMV-related end-organ disease. The patient was maintained on dual foscarnet and letermovir antiviral therapy through day 188. During this time, her CMV levels declined until undetectable, her diarrhea improved, and her GvHD was downgraded significantly (stage I–II). With resolution of her DNAemia and no evidence of ongoing organ-specific disease, the foscarnet was stopped, and the patient was managed on oral letermovir (480 mg once daily) for ongoing secondary prophylaxis until her discharge from the hospital.

After discharge, the patient initially did very well on letermovir monotherapy, but there was concern for potential medication noncompliance despite close interval follow-up. She developed recurrent diarrhea with weight loss, and repeat CMV testing revealed evidence of recurrent DNAemia. Repeat resistance testing on day 257 revealed a UL56 mutation (R369S) that conferred resistance to letermovir, continued UL97 resistance to ganciclovir (A594V), and a mixed population of virus resistant to foscarnet in the UL54 gene (N495K) and wild-type virus. Resistance testing results over time are listed in Table 1.

Table 1.

CMV Resistance Testing Results Over Time

Gene Target and Antiviral Drug	Resistance on:
	Day 48	Day 90	Day 257
UL97
Ganciclovir	None detected	At site A594V	At site A594V
UL54
Cidofovir	None detected	None detected	None detected
Foscarnet	None detected	None detected	At site N495Ka
Ganciclovir	None detected	None detected	None detected
UL56
Letermovir	Not tested	Not tested	At site R369S

Abbreviation: CMV, cytomegalovirus.

^aIn the sample submitted, both a wild-type nucleotide sequence and a nucleotide sequence consistent with CMV antiviral resistance were detected at site N495K, which suggests the presence of a mixed CMV population.

On day 266, 2.5 months after resolution of her initial DNAemia and hospital discharge, the patient was readmitted with fevers, weight loss, and diarrhea with rising CMV DNAemia. She was placed on combination high-dose IV ganciclovir (10 mg/kg every 12 hours) and IV foscarnet (90 mg/kg every 12 hours) antiviral therapy, and her DNAemia decreased.

DISCUSSION

To our knowledge, our case represents the first published use of letermovir for the treatment of resistant CMV infection in a pediatric patient. The first-line agent for CMV treatment is ganciclovir and, if appropriate, the valine ester oral equivalent, valganciclovir. These agents are prodrugs that require phosphorylation by viral serine-threonine kinase and inhibit the viral DNA polymerase to suppress viral replication. Mutations at the UL54 (DNA polymerase) or UL97 (viral phosphotransferase) gene can result in resistance to ganciclovir. Foscarnet works by inhibiting viral DNA polymerase but often causes reversible nephrotoxicity and electrolyte abnormalities [4], and mutations in the UL54 gene can be associated with foscarnet resistance. Cidofovir also acts on the viral polymerase, and select mutations in the UL54 gene can lead to cidofovir resistance. However, cidofovir is associated with a higher risk of nephrotoxicity that can be irreversible [5]. Although other antiviral agents, such as maribavir or brincidofovir, are considered for the treatment of CMV, neither agent is approved by the US Food and Drug Administration, and data showing its definitive efficacy for prevention, suppression, and treatment of CMV infection and disease are lacking.

Despite experiencing an early response to ganciclovir monotherapy, the patient’s increased immunosuppression likely contributed to delayed clearance of CMV DNAemia and subsequent ganciclovir resistance. Given the comparison of potential adverse-effect profiles, foscarnet was selected over cidofovir. Because she had a rising CMV DNAemia level while on foscarnet monotherapy, letermovir was added for salvage therapy despite the limited pediatric data available.

The US Food and Drug Administration approved letermovir in 2017 for CMV prophylaxis in adult stem cell transplant recipients, but the drug is not approved for treatment, and very limited published clinical experience showing its efficacy for the treatment of active infection or disease exists. In addition, limited pediatric pharmacological data for letermovir prophylaxis or treatment exist [5]. Letermovir works by inhibiting the viral terminase complex that subsequently inhibits proper viral DNA cleavage and prevents the packaging of new viral capsids [6, 7]. Although letermovir has a favorable adverse-effect profile compared to that of other conventional anti-CMV agents, studies have found worsening enteritis associated with letermovir use [6]. As was the case for our patient and many HSCT recipients, it was difficult to differentiate letermovir-induced enteritis from progressive GvHD or effects of ongoing CMV colitis.

In addition, in vitro studies have suggested that the threshold for letermovir resistance might be lower [7]. Although the mutation (R369S) that confers letermovir resistance in this case has not been demonstrated clinically, in vitro studies have found rapid mutations to UL56, including codon 369 [8, 9]. A systemic review by Chen et al [10] of randomized trials of prophylactic antiviral agents against CMV after allogeneic HSCT detailed letermovir resistance (UL56 V236M) in a patient enrolled in the Marty et al [6] phase 3 clinical trial with an additional mutation (UL56 C325W) identified in 1 of 48 patients with detectable CMV DNA who began letermovir treatment. However, the exact incidence of letermovir resistance in clinical settings is not well reported. As in our case, after switching from an IV to oral letermovir formulation, her severe intestinal GvHD could have reduced enteric absorption and serum levels. Medication noncompliance also might have led to lower serum levels and contributed to her development of resistance.

This case reveals the potential for development of antiviral resistance within the setting of sustained DNAemia. Many factors, including increased immunosuppression, variable serum letermovir levels, and a low barrier to resistance, likely contributed to the development of letermovir resistance in this patient. Additional studies examining the pharmacokinetic data in pediatric populations are needed, because the ideal dosing and serum concentrations of letermovir are not known. We attempted to obtain serum letermovir levels, but it was not feasible commercially or through the pharmaceutical company. More data are needed to examine the role of letermovir for salvage treatment purposes for pediatric patients with refractory CMV disease and limited treatment options.

Notes

Authorship contributions. J. T. K. conceived the article, wrote the manuscript, and is the primary author of the article. B. B. helped conceive the paper’s design and was a coauthor and editor for the article. M. G. V. analyzed the data and contributed to manuscript preparation. S. P. reviewed the data and helped construct the discussion and Figure 1. V. P. reviewed the data and helped construct the discussion and Figure 1. D. L. helped conceive the paper’s design and was a coauthor and editor for the article. Y.-C. C. helped conceive the paper’s design and was a coauthor and the primary editor for the article.

Financial support. This work was supported by the National Institutes of Health (grant T32-AI007062-40 to J. W. Sleasman [principal investigator] and J. T. K. [fellow trainee]) and the National Cancer Institute (grant T32-CA057726-26 to K. Ribisl [principal investigator] and M. G. Varga [fellow trainee]).

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.