Use of Long-Term Suppressive Acyclovir after Hematopoietic Stem-Cell Transplantation: Impact on Herpes Simplex Virus (HSV) Disease and Drug-Resistant HSV Disease

The widespread use of prophylactic acyclovir (ACV) therapy has raised the concern that such practice may lead to the emergence of ACV-resistant herpes simplex virus (HSV) [1]. Several centers have recently reported cases of clinically significant HSV-1 infections caused by ACV-resistant HSV in hematopoietic cell transplant (HCT) recipients who had prolonged exposure to ACV [2–5]. The purpose of the present study was to evaluate the effect that 2 long-term regimens of ACV, sequentially introduced to prevent varicella-zoster virus (VZV)–reactivation disease after HCT, has on the incidence of ACV-resistant HSV disease.

Patients, materials, and methods. Three cohorts of VZV- and HSV-seropositive autologous and allogeneic T cell–replete HCT recipients were retrospectively analyzed: cohort 1 (HCT during 1996–1999) received ACV at 250 mg/m² iv twice daily until engraftment and then 800 mg twice daily, for 30 days after HCT; cohort 2 (HCT during 1999–2002) received either ACV at 800 mg twice daily or (if steroid dose was ⩾0.5 mg/kd/day) valacyclovir at 500 mg twice daily, until 1 year after HCT); and cohort 3 (HCT during 2002–2003) received the same regimen as did cohort 2, plus extended use (i.e., after 1 year) in patients who were still receiving immunosuppressive drugs at 1 year. Patients signed an informed consent allowing analysis of their clinical information, and the study was approved by the Institutional Review Board at the Fred Hutchinson Cancer Research Center.

Cases of HSV disease were confirmed by either laboratory testing of lesion samples submitted for conventional culture, direct immunofluorescence assay, or polymerase chain reaction using HSV-DNA. In cases of clinical failure of ACV, isolates were tested for drug susceptibility by either the DNA-hybridization method (EC₅₀, ⩾4 µg/mL) or plaque-reduction assay (EC₅₀, ⩾3 µg/mL).

Time-to-event analysis was performed; cumulative-incidence estimates, in which death, second transplantation, and morphological relapse were treated as competing risk events, were generated for both overall HSV disease and ACV-resistant HSV disease [6, pp. 168–9]. The log-rank test was used to compare the cohorts in terms of the underlying hazard of both overall HSV disease and ACV-resistant HSV disease. Cox regression models, in which age (⩾45 years vs. <45 years), cell source (bone marrow vs. peripheral blood), type of donor (autologous vs. allogeneic), conditioning regimen (nonmyeloablative vs. myeloablative), and donors'/recipients' cytomegalovirus serostatus were considered as potential confounders, were used to estimate the risk of HSV disease and of postengraftment neutropenia (evaluated until day 100 after HCT); because of the small number of events in cohort 3, cohorts 2 and 3 were combined for the Cox regression analysis. The proportional-hazard assumption was tested [7] and held for the first 6 months after HCT; subsequently, the hazard ratio (HR) changed (i.e., decreased). The HR for disease at 2 years therefore reflects an average over the entire time period. HRs for both 6 months and 2 years are presented. Because of the small number of cases, adjusted risk estimates for ACV-resistant HSV disease could not be provided. The population fraction of cases of HSV disease prevented by the use of long-term ACV was calculated as described elsewhere [8]. The number of patients necessary for treatment to prevent 1 episode of HSV disease per month was calculated as 1/(no incidence of ACV − incidence of ACV). Given the variation in the amount of follow-up time of the 3 cohorts, this analysis was restricted to the first 2 years after HCT; the analysis was performed by use of Stata (version 9.0).

Results. The characteristics of HSV disease and ACV-resistant disease are shown in table 1. Figure 1A shows the cumulative incidence of HSV disease. The probability of a first instance of HSV disease at year 2 after transplantation decreased from 31.6% (95% CI, 28.0%–35%) in cohort 1 to 3.9% (95% CI, 2.7%–5.2%) in cohort 2 and to 0% in cohort 3 (log rank test, P<.001 for cohort 1 vs. cohort 2 and for cohort 2 vs. cohort 3); in all 3 cohorts, the difference between the strategies occurred mostly during the first year (cohort 1, 30.5% [95% CI, 27.1%–34.0%]; cohort 2, 3.2% [95% CI, 2.2%–4.0%]; cohort 3, 0%); no significant difference was observed during the second year (cohort 2, 0.89% [95% CI, 0.42%–1.69%]; cohort 3, 0%; P=.07).

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Table 1.

Characteristics and outcomes in hematopoietic stem-cell transplant (HCT) recipients in 3 cohorts, which received prophylactic acyclovir (ACV) for 30 days, 1 year, or >1 year

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Figure 1.

Time to first instance of herpes simplex virus (HSV) disease overall (A) and first instance of drug-resistant HSV disease (B), in cohort 1 (30 days of prophylactic acyclovir) and in cohort 2 (1 year of prophylactic acyclovir); no cases of HSV disease occurred in cohort 3.

After adjustment for stem-cell source, conditioning regimen, and donor type, the HR of HSV disease remained significantly lower, both at 6 months (HR, 0.05 [95% CI, 0.03–0.08]; P<.001) and at 2 years (HR, 0.07 [95% CI, 0.05–0.1]; P<.001), in patients who received long-term suppressive prophylactic ACV (i.e., cohorts 2 and 3 combined) than those in patients who received 30 days of prophylactic ACV (i.e., cohort 1).

Year-2 incidence rates (per 1000 person-months) of all episodes of HSV disease were 36 (95% CI, 32–41), 3 (95% CI, 2–4), and 0, in cohorts 1, 2, and 3, respectively.

The cumulative incidence of ACV-resistant HSV disease is shown in figure 1B. The probability of ACV-resistant HSV disease at 2 years after HCT decreased from 1.3% (95% CI, 0.8%–2.7%) in cohort 1 to 0.22% (95% CI, 0%–0.8%) in cohort 2 (log-rank test; P=.006) to 0% in cohort 3 (log-rank test, cohort 2 vs. 3; P=.3). The year-2 incidence rates (per 1000 person-months) of all episodes of ACV-resistant HSV disease were 0.9 (95% CI, 0.5–1.9), 0.1 (95% CI, 0.04–0.5), and 0, in cohorts 1, 2, and 3, respectively. Hematological toxicity, evaluated on the basis of the development of postengraftment neutropenia up to 100 days after HCT (the period for which complete hematology records are available), did not differ between cohorts (table 1).

Discussion. The present study shows that long-term administration of ACV for prevention of VZV reactivation after HCT is associated with a significant reduction of both overall HSV disease and ACV-resistant HSV disease, at 2 years after HCT. The protective effect was predominantly seen during the first year after HCT. The improved prevention outcome for cohort 3, compared with that for cohort 2, during the first year after HCT is likely due to better adherence; alternatively, this effect may be due to other changes in practice, which may have led to a reduced risk; however, although the latter possibility cannot be ruled out, major factors such as the use of peripheral-blood cells did not differ between cohort 2 and cohort 3. The present study's findings are consistent with those of an earlier report, which suggested that ganciclovir used for prevention of cytomegalovirus is associated with a lower risk of ACV-resistant HSV disease [2].

The present study's findings contrast with the increasing incidence of ACV-resistant HSV disease observed in several HCT centers [2, 3]. The differences between our center and others, both in transplantation procedures and in the doses of prophylactic ACV used, may account for the observed discrepancies: First, the profound immunosuppression caused by the use of T cell–depleted HCT performed in other centers [2, 3] may increase the emergence of ACV-resistant HSV disease. Mutant (i.e., ACV-resistant) HSV may be present in any lesion and, because of their mild pathogenicity, are easily controlled by T cell immunity [9]; it is therefore probable that replication of the mutant viruses is enhanced during periods of profound T cell immunodeficiency. Second, the higher prophylactic doses—either 1600 mg ACV/day or 1000 mg valacyclovir/day—used in the present study, compared with the ⩽1000-mg-ACV/day dose used by other transplantation centers [2, 3], may be crucial in the abrogation of HSV reactivation and, consequently, in the emergence of mutant HSV after HCT. Therefore, if non–T cell–depleted recipients require either ⩾800 mg of ACV or 500 mg of valacyclovir twice daily to prevent the emergence of ACV-resistant HSV disease, even higher doses might be necessary in the treatment of T cell–depleted HCT recipients, as has been suggested in a recent study [3].

Although the present study provides strong evidence that prophylactic ACV given for 1 year prevents most cases of HSV disease, the benefit obtained from extending it for >1 year is less clear, because of the very low incidence of HSV disease observed after 1 year. It is theoretically conceivable that HSV cases may occur even after 2 years; however, the fact that none of the patients in the present study developed HSV disease during the second year after 1 entire year of acyclovir (cohort 2) is a significant finding, because one would expect to see an increase in the incidence of HSV disease soon after discontinuation of medication, as has been described for VZV disease treated with prophylactic ACV [10]). Therefore, although the occurrence of very late cases (i.e., those occurring after ⩾2 years) cannot be ruled out, they are likely to be very rare and sporadic.

The present study was limited by the use of a noncontemporaneous control group. However, results for HSV disease remained significant after adjustment for several variables related to changing transplantation procedures, including the use of peripheral blood as the cell source. Also, in a prospective randomized trial of transplantation, HSV incidence when bone marrow was the cell source was not different from that when peripheral blood was the cell source [11]. Because of the small number of events, multivariable modeling could not be performed for drug-resistant HSV disease; however, we believe that improved and more-accessible viral diagnosis over time has made identification of cases of drug-resistant HSV disease more likely in recent years. Also, the concern for antiviral resistance has increased, rather than decreased, since the introduction of prophylactic therapy using ACV. Indeed, in light of the increased use of ACV, we were very concerned with drug resistance, and virtually every suspected case was analyzed by laboratory testing. Because drug-resistant HSV disease would be treated differently, the recommendation to analyze suspicious lesions was also forwarded to patients' physicians after the patients were discharged from our center. Therefore, we do not believe that the absence of ACV-resistant HSV disease is due to a lack of testing. Finally, the evaluation of the safety of our strategy was limited by the retrospective nature of the study. We did not find differences between the groups for hematologic toxicity during the first 100 days after HCT. A previous randomized placebo-controlled study did not show any toxicities when the same regimen was administered for 1 year [10]. Furthermore, no major toxicities have been reported with even higher doses of either acyclovir (e.g., 3.2 g/day) or valacyclovir (e.g., 8 g/day), compared with the doses used in our study [12, 13]. The safety and tolerability of acyclovir therapy for >1 year has not been studied. In conclusion, the present population-based study suggests that, contrary to earlier predictions [2–5], the introduction of widespread use of prophylactic ACV at the dose used in these 3 cohorts has not led to an emergence of drug-resistant HSV disease but, rather, appears to have completely eliminated the problem during the first 2 years after transplantation.

• Received January 16, 2007.
• Accepted February 22, 2007.

• © 2007 by the Infectious Diseases Society of America