Clinical and In Vitro Evaluation of Cidofovir for Treatment of Adenovirus Infection in Pediatric Hematopoietic Stem Cell Transplant Recipients

Post—hematopoietic stem cell transplantation (HSCT) adenovirus infections were identified in 31 of 204 consecutive pediatric HSCT patients, 18 of whom had severe manifestations of infection. Cidofovir treatment led to clinical improvement in 8 of 10 patients with severe infection and to virologic clearance in 9 patients. In vitro susceptibility to cidofovir was demonstrated in 12 clinical adenovirus isolates. Cidofovir is a promising treatment option for this population.

Adenoviruses are serious pathogens among immunocompromised patients [1, 2], with an incidence of infection of up to 20% among hematopoietic stem cell transplant (HSCT) recipients [3] and case-fatality rates that approach 60% [4]. Respiratory, gastrointestinal, and genitourinary infections with adenovirus are common in children undergoing HSCT [5, 6], and disseminated illness is often fulminant and fatal. Cidofovir has in vitro activity against adenoviruses [7–9], and several case studies [3, 10–14] report successful use of cidofovir for treatment of adenovirus infection in HSCT and solid organ transplant recipients. We reviewed the findings regarding adenovirus infection in a pediatric HSCT population and describe our experience with cidofovir therapy.

Methods. We reviewed data from a database on 204 consecutive patients who underwent HSCT at our institution during the period of 1 January 1994 and 31 March 2003, and we analyzed records for patients with adenovirus infection for a minimum of 200 days after transplantation, because data for each patient were available for at least this duration at the start of the study. Testing for adenovirus infection was performed only for symptomatic patients; there was no routine surveillance. Diagnosis was confirmed by detection of adenovirus antigen in respiratory specimens by immunofluorescence, by observation of typical adenovirus particles in stool specimens concentrated by ultracentrifugation, and by conventional and shell-vial centrifugation culture of nasal secretions, bronchoalveolar lavage specimens, urine samples, or stool samples, with use of monoclonal antibodies and immunofluorescence for confirmation. Serotyping of adenovirus isolates was not done. Patients with adenovirus detected from >1 site during the same illness were classified as having disseminated infection.

Patients with adenovirus infection were categorized as having severe infection either if they had disseminated infection or if infection at 1 site was accompanied by severe clinical manifestations attributable to the infection [12]; all other infections were categorized as mild. Resolution of infection was defined as clinical improvement and as virologic evidence of clearance; patients with intermittent positive test results after having symptomatic illness were considered to have continued infection until specimens yielded consistently negative test results with continued absence of symptoms over a 30-day period. Recurrence of infection was defined as evidence of adenovirus at any site >30 days after resolution of prior infection.

Patients treated with cidofovir received 5 mg/kg intravenously once weekly for up to 6 weeks, then once every 2 weeks for up to 3 additional doses. Treatment was discontinued on the basis of intolerance, clinical improvement, and negative adenovirus test results. Probenecid and intravenous hydration were given before and after cidofovir infusion, in accordance with the package insert [15]. Cyclosporine doses were not given on the day of cidofovir infusion, and prophylactic acyclovir was not given for the 4 days following infusion. In some patients, cidofovir doses were held or reduced because of renal dysfunction.

The antiviral susceptibility of randomly selected clinical isolates was determined by plaque reduction. Confluent monolayers of A549 cells in 12-well plates were inoculated with 50–100 plaque-forming units of virus isolates in 0.4 mL of maintenance medium. After adsorption for 45 min at 37°C, 2 mL of 0.4% agarose overlay containing serial concentrations of antiviral agents was added. After incubation for 5 and 10 days, 1 mL of 0.4% agarose overlay was added, with sufficient drug for each well to keep the drug concentrations constant. Plates were monitored daily for up to 17 days for cytopathologic signs of adenovirus. Monolayers were then washed, fixed, and stained with crystal violet. Plaques were counted in triplicate wells, and the drug concentration required to reduce plaque number by 50% (IC₅₀) was calculated.

Cytotoxic effects of antiviral agents were examined by measuring cell DNA synthesis. A549 cells were plated in 96-well plates with various drug concentrations. After 3 days, [methyl-³H]-thymidine was added for a 6-h incubation at 37°C. Cells from each well were captured on glass fiber filters, washed 10 times with 10 mL of water and once with 10 mL of cold ethanol, dried at room temperature, and assayed for [methyl-³H]-thymidine incorporation. The drug concentration required to reduce [methyl-³H]-thymidine incorporation by 50% (CC₅₀) was determined. The selectivity index was defined as the ratio of CC₅₀ to IC₅₀.

Life table or Cox proportional hazard regression models were used to analyze associations between study variables [16]. Logistic regression or Fisher's exact test was used to examine associations between outcome variables and risk factors. The Wilcoxon rank sum test was used for comparison of data that were not normally distributed. All statistical analyses were conducted using SAS software, version 8 (SAS Institute) [17].

Results. Table 1 shows risk factors for adenovirus infection in our HSCT population. Infection was more likely in allogeneic HSCT recipients than in autologous transplant recipients. Infection in allogeneic HSCT recipients was more likely to occur when the transplants were from unrelated or related HLA-mismatched donors, compared with transplants from related HLA-matched donors. The median time to onset of infection after HSCT was 53 days, with 84% of infections occurring before day 200. Most initial adenovirus infections were in the gastrointestinal tract (71%). Disseminated infection occurred in 11 of 31 patients.

Adenovirus infections were severe in 58% of patients who developed infection, all of whom underwent allogeneic HSCT. Severe infection was more common among recipients of transplants from related HLA-mismatched donors or from unrelated donors; however, these trends were not statistically significant.

The mortality rate during the study period was 32% for adenovirus-infected HSCT recipients, which was similar to that for HSCT recipients without infection (P =.17). Among allogeneic HSCT recipients with severe infection, 9 deaths occurred (50% mortality), with 5 patients having evidence of adenovirus infection (by culture and/or PCR) at the time of their deaths. The mortality rate among HSCT recipients with severe infection was not statistically different from the rate among those with mild infection.

Ten HSCT recipients with adenovirus infection were treated with cidofovir (table 2). All of these patients had severe infection at the time of treatment. Clinical resolution of infectious symptoms after receipt of treatment with cidofovir occurred in 8 patients, and virologic clearance was observed in 9 patients. Recurrence of infection before the end of the follow-up period occurred in 5 previously treated patients. Recurrent infection was severe in 1 patient, who required additional cidofovir therapy. All patients treated with cidofovir who had recurrence had subsequent clinical and virologic resolution of infection. The toxicity of treatment was assessed by changes in the serum creatinine level, with 3 patients showing increases of 50% above the baseline level.

Of 8 patients with severe infection who received supportive care only, clinical resolution of infection was observed in 4 patients, and virologic resolution was observed in 5 patients (table 2). One patient with virologic resolution of gastrointestinal infection had persistent symptoms attributable to graft-versus-host disease. Three of 8 patients who were treated with only supportive care died without resolution of initial infection, compared with 1 of 10 cidofovir recipients (table 3). One of 2 patients with recurrent infection who was treated with only supportive care also died without resolution of adenovirus infection. Patients treated with cidofovir for severe adenovirus infection had higher survival rates, compared with those receiving supportive care only (OR, 0.31). However, this difference did not attain statistical significance (P =.11).

All clinical isolates tested for drug susceptibility were susceptible to cidofovir at concentrations well below the CC₅₀ of the human cells tested (table 4). The CC₅₀ for cidofovir against A549 cells was 9.69 µg/mL, which is at least 2 times the highest IC₅₀ determined and is generally ⩾4-fold higher. Four isolates were also tested for susceptibility to IFN-α, IFN-γ, and ribavirin. IFN-α and/or IFN-γ reduced plaque formation, compared with the control, by up to 46%; at these levels, IFNs were cytotoxic. The combination of IFN-α and IFN-γ was not found to have synergy for adenovirus plaque formation in vitro (data not shown). The IC₅₀ of ribavirin, tested at concentrations of 50–200 µg/mL, was 100–150 µg/mL. The CC₅₀ for ribavirin against A549 cells was 9.86 µg/mL, which was significantly lower than the levels necessary for reduction of viral plaque formation, indicating a poor selectivity index.

Discussion. This report describes symptomatic adenovirus infections in 31 children after HSCT, including 10 severe infections treated with cidofovir. The overall infection rate and incidence of severe infection are comparable to those reported in prior publications [1, 3, 5, 10, 11, 18–21], as are risk factors for severe adenovirus infection.

The mortality rate for HSCT recipients with adenovirus infection has been reported to range from 12% to >60% [10, 19], with higher mortality associated with severe adenovirus infection. The current study found that overall survival rates for adenovirus-infected allogeneic HSCT recipients were not different from those for the adenovirus-uninfected population, suggesting that management strategies for adenovirus infection in our patient population may have favorably influenced their outcome.

Cidofovir treatment was offered to and used by 10 patients, although patients were not randomized to receive treatment. However, all had severe disease and were generally more ill than were those who were treated with supportive care. Only 1 death was noted in the treatment group before resolution of adenovirus infection, whereas 3 of 8 patients with severe infection who received only supportive care died with active infection. Prior reports [3, 10–13] have also described success using cidofovir treatment in the HSCT population.

Adenovirus isolates recovered from HSCT recipients evaluated for in vitro efficacy of cidofovir showed a favorable selectivity index, whereas neither IFN-α nor IFN-γ significantly inhibited adenovirus growth at levels achievable in humans. Additionally, adenovirus was inhibited by ribavirin only at levels that are toxic to A549 cells in vitro and higher than levels that are generally achieved in serum with systemic treatment (∼4 µmol/L [22]). Ribavirin also did not have demonstrated clinical benefit in a recent prospective study of adenovirus infection in 4 HSCT recipients [23]. The susceptibility of our adenovirus isolates to cidofovir was at levels similar to those reported in previous studies [9, 24]. Naesens et al. [8] also recently found in vitro efficacy for cidofovir against a clinical adenovirus isolate and found no activity for ribavirin.

Our study provides additional support for an efficacy trial of cidofovir for severe adenovirus infection in relation to allogeneic HSCT, and it further demonstrates that cidofovir can be safely used in these severely ill children. Finally, most adenovirus isolates in our study were susceptible to cidofovir, as determined using a conventional in vitro assay. Current studies concerning the pharmacokinetics of cidofovir in pediatric HSCT patients (R. Giller, personal communication) will provide relevant dosing information. An efficacy trial would need to be randomized but could focus on preemptive treatment of early infection.

Acknowledgments

Financial support. Louis and Sidelle Bruckner Memorial Fund

Potential conflicts of interest. All authors: no conflicts.

Figures and Tables

Table 1

Risk for adenovirus infection among populations of hematopoietic stem cell transplant (HSCT) recipients.

Table 2

Treatment patients and outcomes for cidofovir treatment, compared with supportive care alone.

Table 3

Characteristics of patients with severe adenovirus infection who died without resolution of infection.

Table 4

In vitro efficacy of cidofovir in clinical isolates of adenovirus recovered from hematopoietic stem cell transplant (HSCT) recipients.