Treatment of Adult Varicella with Oral Acyclovir A Randomized, Placebo-controlled Trial Mark R. Wallace, MD; William A. Bowler, MD; Nancy i. Murray, MD; Stephanie K. Brodine, MD; and Edward C. Oldfield III, MD
• Objective: To assess the efficacy of oral acyclovir in treating adults with varicella and to describe the natural history of adult varicella. • Design: Double-blind, placebo-controlled randomized trial. • Setting: A naval hospital. • Patients: One hundred forty-eight of 206 consecutive adult active duty Navy and Marine Corps personnel who were hospitalized for isolation and inpatient therapy of varicella and who could be treated within 72 hours of rash onset completed the study. The diagnosis of varicella was confirmed by acute and convalescent serology in 143 of 144 patients with available paired sera. • Intervention: Patients were randomly assigned to receive either acyclovir, 800 mg orally five times per day for 7 days, or an identical placebo. Separate randomization codes were used for patients presenting within 24 hours of rash onset and for those presenting 25 to 72 hours after rash onset. • Measurements: Daily lesion counts, symptom scores, temperature measurements, and laboratory tests were used to monitor the course of the illness. • Results: Early treatment (initiated within 24 hours of rash onset) reduced the total time to (100%) crusting from 7.4 to 5.6 days (P = 0.001) and reduced the maximum number of lesions by 46% (P = 0.04). Duration of fever and severity of symptoms were also reduced by early therapy. Late therapy (25 to 72 hours after rash onset) had no effect on the course of illness. Only four patients had pneumonia, and no encephalitis or mortality was noted. • Conclusions: Early therapy with oral acyclovir decreases the time to cutaneous healing of adult varicella, decreases the duration of fever, and lessens symptoms. Initiation of therapy after the first day of illness is of no value in uncomplicated cases of adult varicella. The low frequency of serious complications of varicella (pneumonia, encephalitis, or death) precluded any evaluation of the possible effect of acyclovir on these outcomes.
Varicella (chickenpox) is typically a benign illness of childhood characterized by a vesicular rash and constitutional symptoms. Although only 1% to 2% of varicella cases occur in adults, the illness is often more severe than in children and is associated with a prolonged course and a 25-fold increased risk for mortality (1-4). Safe and effective therapy might decrease the morbidity of adult varicella and could prove cost effective by returning students and workers to their usual activities more rapidly. Acyclovir, an acyclic guanosine analog, is an antiviral agent with a proven safety record and established activity against the varicella zoster virus (VZV) (5). Blood levels sufficient to inhibit replication of varicella zoster virus are achieved with intravenous acyclovir; levels achieved after oral administration are much lower and may not exceed the in-vitro inhibitory concentration for varicella (6). Several trials have shown the value of intravenous acyclovir in treating varicella zoster virus infections in the immunocompromised host (7-10). Intravenous acyclovir has also been shown to shorten the course of varicella in normal adults (11), but the expense and potential hazards of parenteral therapy can rarely be justified. Trials of oral acyclovir in immunocompetent children with varicella have recently been completed. Fever and duration of illness were reduced when therapy was started within the first 24 hours of illness (12, 13). The results of two small, uncontrolled trials of oral acyclovir in normal adults with varicella have also suggested a benefit (14, 15), but no randomized trial of oral acyclovir in immunocompetent adults has been done. We conducted a double-blind, placebo-controlled trial of oral acyclovir in the treatment of adult varicella. Our goals were to assess the value of therapy, to determine how long after rash onset therapy remained beneficial, and to describe the natural history of adult varicella. Methods Patients Adults (17 years of age or older) who had developed a characteristic varicella rash within the previous 72 hours and who were admitted to the Naval Hospital, San Diego, were eligible for enrollment. Servicemen and women with varicella are routinely hospitalized if they reside on board a ship or in a barracks; this is done both to control infection and to facili-
Annals of Internal Medicine. 1992;117:358-363. From the United States Naval Hospital, San Diego, California. For the current author addresses, see end of text. 358
Drug Generic Name acylovir
Brand Name Zovirax
1 September 1992 • Annals of Internal Medicine • Volume 117 • Number 5
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19758/ by a University of California San Diego User on 03/05/2017
tate patient care. Patients excluded from study participation included those who were pregnant, had received any antiviral therapy during the previous week, had been vaccinated for varicella, were positive for human immunodeficiency virus (HIV), or had received any immunoglobulin products within the previous week. A history of either renal disease or acyclovir intolerance was also a criterion for exclusion. A complete medical history and a physical examination were done at study entry. Written informed consent was obtained from all patients. The study was approved by the hospital's committee for the Protection of Human Subjects. Treatment Regimens Patients were randomly assigned in groups of eight to receive either acyclovir, 800 mg orally five times per day (at 0600, 1000, 1400, 1800, 2200 hours) for 7 days or an identical placebo. All 35 doses were administered by nursing personnel. Two separate randomization codes were used: one for patients admitted within 24 hours of rash onset (early group) and one for patients admitted 25 to 72 hours after rash onset (late group). The patient's best estimate of rash onset was used; prodromal symptoms of headache or fever were not used for grouping patients. Evaluation of Disease Status Clinical evaluations were done at study entry and daily thereafter. The number of lesions within a standard 25x25 cm area (defined as the "pox box") on the chest and back of each patient was determined, and all new lesions were noted. The "pox box" was used to facilitate accurate counting. Because the upper torso often exhibits acneiform lesions, which can be mistaken for late varicella lesions, this area was not used for study evaluation. The percentage of lesions that were maculopapular (erythematous bump without fluid), vesicular (blisterfluid filled or collapsed), crusted, or healed were estimated to the nearest 10 percentage points. Counting and grading of lesions were done daily through day 7, or until all lesions were crusted or healed. Temperature, blood pressure, pulse, and respiratory rate were recorded every 8 hours by nursing personnel. Patients were asked daily about pruritus, anorexia, and lethargy. A quantitative measure of constitutional illness was derived by summing daily the severity scores for anorexia, lethargy, and fever. These complaints were graded as none (0 points), mild (1 point), moderate (2 points), or severe (3 points). Conversion of fever scores to 0, 1, 2, or 3 points corresponded to daily recorded temperatures of less than 37.8 °C, 37.8 to 38.3 °C, 38.4 to 39.4 °C, and 39.4 °C or greater. Laboratory studies on the day of entry assessed the leukocyte count, hematocrit, and platelet count, as well as levels of alanine aminotransferase (ALT), aspartate aminotransferase (AST), total bilirubin, alkaline phosphatase, blood urea nitro-
gen, and serum creatinine. Serum samples were also evaluated for acute varicella serology. Liver function tests were repeated on day 4, and all laboratory tests were repeated on day 7. Convalescent varicella serum was obtained on day 21 (days 14 through 28 were acceptable); all sera was frozen at -70 °C. A chest radiograph was done on day 1. If radiographic results showed an infiltrate, the test was repeated weekly until results were normal. If any laboratory value (except hematocrit) was abnormal on day 7, tests were repeated weekly until normal. Diphenhydramine (50-mg tablets) and acetaminophen (325mg tablets) were available to patients on request to treat pruritus and constitutional symptoms, respectively. Acetaminophen was given every 4 hours to patients whose temperatures were over 38.5 °C. Patients were hospitalized a minimum of 7 days and discharged when all skin lesions were crusted or healed. Serologic Tests Immunoglobulin G (IgG) antibodies to varicella zoster virus were measured by enzyme immunoassay. The wells of microtiter plates (Immulon 2, Dynatech Laboratories, Chantilly, Virginia) were coated with varicella zoster virus antigen or uninfected melanoma cell control and incubated overnight at 4 °C. The plates were washed, and 3% bovine serum albumin in phosphate buffered saline was added to block nonspecific binding. Each test serum was added to duplicate wells and serially diluted from 1:4 to 1:65 536. Biotinylated anti-IgG was added, amplified with an avidin-biotin complex and color change was detected by adding o-phenylene-diamine dihydrochloride and H 2 0 2 (Vectastain ABC Kit, Vector, Burlingame, California). Antibody titers equal to or greater than 1:4 were considered positive. Statistical Analysis The Kaplan-Meier product-limit method was used to estimate all times to events, calculated from the date of study entry to the date of the event. Comparisons of time differences between the treatment groups were done using the log-rank test. Treatment differences were assessed using two-sided tests, and statistical significance was set at P = 0.05. Results were reported separately for patients with rash onset occurring 24 hours or less before study entry and those with onset occurring more than 24 hours before study entry. Ninety-five percent CIs were computed for the relative risk of acyclovir. The relative risk of acyclovir was used to compare the response rate of patients receiving acyclovir with that of patients receiving placebo. Relative risks less than 1.00 indicated that acyclovir performed better than placebo. Differences in medians between treatment groups were compared using the Wilcoxon rank-sum test. Ninety-five percent CIs for the difference are given where appropriate. Percentages of patients in
Table 1. Patient Characteristics at Study Entry Characteristic
Age, y Mean (±SD) Race, n(%) Black White Filipino Other Lesions, n* Mean (±SD) Duration of rash, h Mean (±SD) Temperature at study entry, °C Mean (±SD)
Early Group (< 24 Hours) Acyclovir Placebo in = 38) (n = 38)
Late Group (> 24 to 72 Hours) Placebo Acyclovir (n = 36) (n = 36)
20.26 (2.97)
20.00 (2.51)
21.00 (2.97)
19.67 (2.38)
14 (37) 20 (53) 3(8) 1(3)
13 (34) 17 (45) 5(13) 3(8)
10 (28) 17 (47) 7(19) 2(6)
13 (36) 13 (36) 8(22) 2(6)
193 (170)
199 (220)
248 (226)
198 (187)
17.2 (6.2)
18.8 (5.5)
44.3 (11.64)
45.4 (11.34)
37.8 (1.6)
37.4 (1.5)
37.7 (1.6)
37.6 (1.5)
* Within the "pox box" area; chest and back. See text for description. 1 September 1992 • Annals of Internal Medicine • Volume 117 • Number 5
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19758/ by a University of California San Diego User on 03/05/2017
359
Table 2 . Cutaneous Effects of Varicella Variable
Time to maximum number of skin lesions, d Time of new lesion formation, d Time to onset of cutaneous healing, d Time to 100% crusting, d Maximum number of lesions, n*
Early Group (< 24 Hours) Acyclovir Placebo P Value (n = 38) (n = 38) 1.5 2.7 2.6 5.6 268
2.1 3.3 3.3 7.4 500
0.002 0.03 < 0.001 0.001 0.04
Late Group (25 to 72 Hours) Acyclovir Placebo P Value {n = 36) (n = 36) 1.3 3.0 2.4 7.0 233
1.2 2.3 2.3 6.8 158
> 0.2 0.03 > 0.2 > 0.2 0.03
* Wilcoxon rank-sum test. All other comparisons by log-rank test with median values from Kaplan-Meier product-limit plots.
each treatment group were compared using the Fisher exact test. Results Study Patients All 206 inpatients with varicella who were admitted between 15 February 1989 and 15 March 1990 were evaluated for enrollment. Fifty-three patients were excluded: 44 had rash onset more than 72 hours before hospitalization, 8 refused to participate, and 1 was pregnant. Of the remaining patients, four were withdrawn after 1 to 2 days because of incorrect diagnosis (all had vaccinia secondary to recent smallpox vaccination), and one was withdrawn after developing hives on day 5 of therapy. These 5 patients were excluded from all analyses, leaving 148 evaluable patients. Seventy-six patients were admitted within 24 hours of rash onset (early group), and 72 patients were admitted 25 to 72 hours after rash onset (late group). Only three patients were women. Most of our patients were Navy or Marine recruits undergoing basic training (women recruits receive basic training elsewhere). Characteristics of the evaluable patients are shown in Table 1. The mean age of study patients was 20 years (range, 17 to 33 years). The placebo and treatment groups and the early and late subgroups were similar with respect to initial lesion counts, duration of rash before study entry, and temperature at the time of hospitalization.
Cutaneous Effects Five parameters of cutaneous disease progression were evaluated. By all these criteria, early acyclovir treatment was beneficial, whereas delayed therapy (25 to 72 hours after rash onset) had no positive effect. Early treatment shortened the time to maximum number of lesions from 2.1 to 1.5 days (P = 0.002; relative risk, 0.60; 95% CI, 0.37 to 0.97) and reduced the days of new lesion formation from 3.3 to 2.7 days (P = 0.03; relative risk, 0.70; 95% CI, 0.44 to 1.11). In addition, early acyclovir therapy sped the onset of cutaneous healing (defined as the first day the total number of lesions decreased by 5% and continued to decline thereafter) from 3.3 to 2.6 days (P < 0.001; relative risk, 0.54; 95% CI, 0.33 to 0.88), and produced full crusting in 5.6 compared with 7.4 days (P = 0.001; relative risk, 0.56; 95% CI, 0.34 to 0.93). The maximum number of lesions was also reduced by 46% in the early group (P = 0.04). When the late group was considered, acyclovir had no beneficial effect on any parameter (Table 2). The proportion of patients requiring more than 7 days to achieve 100% crusting was significantly reduced in those patients who were treated within 24 hours. Of 38 patients receiving acyclovir, 12 (38%) took more than 7 days to crust, compared with 22 of 38 patients (58%) receiving placebo (P = 0.02 by chi square). No such benefit was observed in the late group. Effect on Fever For the group enrolled within 24 hours of rash onset, acyclovir treatment reduced the duration of fever from 2.74 to 2.24 days (P = 0.02), (Figure 1). No treatment effect was observed in the late group (2.1 compared with 2.04 days, P = 0.65). Six of 38 patients (16%) treated with acyclovir within 24 hours were febrile for more than 72 hours, whereas 17 of 38 early placebo recipients (45%) were febrile for more than 72 hours (P = 0.02, chi square). Effect on Symptoms and Symptomatic Therapy
Figure 1. Proportion of patients with fever in the early treatment group, acyclovir compared with placebo.
360
Pruritus tended to be less severe and to require less diphenhydramine among patients treated with acyclovir in the early group. This trend was significant only on day 5 (P = 0.025, Fisher exact test); late treatment had no effect. In addition, acetaminophen consumption was consistently less frequent in the early acyclovir treat-
1 September 1992 • Annals of Internal Medicine • Volume 117 • Number 5
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19758/ by a University of California San Diego User on 03/05/2017
ment group. This difference was statistically significant only on days 4 and 5 (P - 0.04 and 0.03, respectively; Fisher exact test). By day 5, only 2.6% of the early acyclovir group was using acetaminophen compared with 18% of the early placebo group. The mean constitutional illness scores peaked on the second day and decreased thereafter for all patients. Although the acyclovir-treated patients had higher mean admission scores, their symptoms diminished more rapidly than did those of the placebo treated groups. This effect was especially pronounced in the early group (Figure 2).
Toxicity One patient who was receiving acyclovir developed urticaria on day 4 and was removed from the study. Another patient who was receiving acyclovir complained of transient nausea but did not require dose modification. Renal abnormalities were not observed. All patients had normal blood urea nitrogen and creatinine levels on both days 1 and 7, except one placebo patient who had a mildly elevated blood urea nitrogen at the time of hospitalization; this abnormality resolved by day 7. Serology
Complications Pulmonary infiltrates were seen on the admission chest radiographs of four patients from the late group, two patients each in the acyclovir and placebo groups. All infiltrates resolved during follow-up. No patient hospitalized with a normal chest radiograph developed pulmonary complaints of sufficient severity to warrant a repeat procedure. Late complications of varicella were infrequent. Two placebo recipients developed secondary skin infections severe enough to require antibiotic therapy. No meningeal or encephalitic manifestations were noted in any patient. Asymptomatic liver function abnormalities were common, and no difference in the frequency, severity, or duration of asymptomatic hepatitis was seen between patients in the acyclovir and placebo groups. Fifty percent of all patients had elevated ALT or AST levels on day 1, 4, or 7. Elevations were generally highest on day 4 or 7 and resolved in 1 to 5 weeks. One patient's ALT and AST levels peaked at 14 times the upper limit of normal on day 7 and slowly decreased to normal over 5 months; an extensive investigation disclosed no other etiology for his prolonged asymptomatic hepatitis. Our patients' hepatic abnormalities were similar to other viral hepatitides in that the ALT level usually exceeded that of the AST. In only five patients did either the ALT or AST level exceed 5 times the upper limit of normal. Slight elevations of total bilirubin were seen in four patients, three of whom had transaminase elevations; none was jaundiced. Alkaline phosphatase levels were mildly increased (1.0 to 1.5 times the upper limit of normal) in three patients. The frequency, duration, and severity of the varicella-associated anemia, thrombocytopenia, and leukopenia were unaffected by acyclovir therapy. Mild thrombocytopenia (platelet count 100 to 150 x 109/L) occurred in 10 of 148 patients (6.8%). Thrombocytopenia resolved before hospital discharge in all cases and was unassociated with bleeding or petechiae. Forty-one patients (27.7%) had a leukocyte count of less than 4.8 x 109/L at the time of hospitalization; the lowest count was 2.5 x 10 9 /L. All cases of leukopenia resolved within 1 week without sequelae. Mild anemia (hematocrit, 0.35 to 0.41 for men) occurred in 30 of 148 patients (20.3%); in only 4 of these patients was the hematocrit less than 0.38.
Acute and convalescent sera were available from 144 of the 148 patients who completed the study. Seroconversion, defined as a two-dilution increase in titer, was documented in 143 of these patients. One patient who was hospitalized with a 3-day history of rash had a high but unchanging titer (1:1024). There was no significant difference in the mean convalescent titers of the acyclovir and placebo groups. Discussion Adult varicella, although uncommon, remains a serious problem and accounts for one fourth of all varicellarelated fatalities in the United States (1, 2). Oral acyclovir therapy reduced the duration of varicella in immunocompetent children when therapy was begun within 24 hours of rash onset (12, 13), but its potential benefit in adults has been unproven. We showed that early acyclovir therapy significantly affected the clinical course of varicella. Time to 100% crusting was reduced by nearly 2 days, and days of fever were reduced by one half day. The reduction in time to 100% crusting was clinically significant in that these sailors and marines were able to return to full duty earlier. All four secondary parameters of cutaneous healing (maximum number of skin lesions, time to onset of cutaneous healing, days of new lesion formation, and days to maximum number of lesions) were also significantly improved by early acyclovir therapy. Significant treatment differences were observed in the
Figure 2. Daily constitutional illness scores of patients treated within 24 hours.
1 September 1992 • Annals of Internal Medicine • Volume 117 • Number 5
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19758/ by a University of California San Diego User on 03/05/2017
361
resolution of constitutional symptoms and signs associated with varicella in the acyclovir-treated patients in the early group. No benefits in reduction of fever, cutaneous involvement, symptoms, or duration of illness were observed in the late group. The results of our study are consistent with those of the acyclovir treatment studies recently completed in normal children (12, 13), which found a similar benefit of acyclovir when started within 24 hours of rash onset, and show that later initiation of therapy is not useful. One explanation for this lack of treatment utility after the first 24 hours of rash onset may be the pattern of viremia in varicella, which peaks just before onset of rash in immunocompetent persons (16). Acyclovir treatment later than 24 hours after onset may be superfluous because viral replication is already being rapidly controlled by the intact host immune system. In immunosuppressed patients with varicella zoster virus infections, however, later initiation of therapy (more than 72 hours after symptom onset) may be of value (10). These patients have presumably sustained viral replication and dissemination related to their immune deficits, allowing acyclovir therapy to yield a treatment benefit much later in the course of their illness. We also observed that patients in the placebo group who presented on day 1 had a more protracted illness than did those who presented on days 2 and 3. Perhaps the patients with more serious disease seek evaluation earlier and thus have more to gain from early intervention. Four patients (2.7%) had radiographically confirmed varicella pneumonia at the time of hospitalization. No patient developed signs or symptoms of pneumonia more than 3 days after rash onset. This result is consistent with those of a previous review, which found 70% of cases of varicella pneumonia occurred within the first 3 days of illness (17). The relatively low incidence of varicella pneumonia in our patients is consistent with the 4.5% found in a recent 485-patient retrospective review of varicella (18). A 1967 study of U.S. Navy personnel found a 16.3% incidence of radiographically diagnosed pneumonia (19); other series have found the incidence of varicella pneumonia in adults with varicella to be as great as 50% (20). In our study, all four patients with pneumonia had a dry cough, but none was dyspneic or hypoxic. Their clinical course was benign, and the radiographic abnormalities resolved within 2 to 3 weeks without sequelae. Although most patients with varicella pneumonia are asymptomatic or have only a dry cough and recover uneventfully, severe cases can occur. Fatal varicella pneumonia has been reported in both immunocompetent (21) and immunocompromised hosts (2) as well as in pregnant women (20). Pneumonia represents the most frequent severe complication of adult varicella. Treatment with acyclovir has not been prospectively studied; however, limited information suggests a benefit (22). We recommend intravenous acyclovir for patients with progressive or symptomatic pneumonia; asymptomatic patients (or those with only a mild cough) should receive oral acyclovir if they present within the first day of their illness. Half of our patients had asymptomatic hepatitis as shown by elevation of either their ALT or AST levels. 362
This is consistent with previous natural history data that showed a 77% incidence of "subclinical hepatitis" (23). Routine testing for hepatic dysfunction in the immunocompetent host with varicella is probably not indicated. The discovery of mild, asymptomatic hepatitis is not unexpected and deserves no particular follow-up or therapy. Acyclovir-related adverse events were rare. One patient developed hives and was dropped from the study. Although rashes have previously been reported among patients receiving acyclovir (5), they have been infrequent. Renal toxicity was not observed, and gastrointestinal complaints were equally infrequent in both the treated and placebo groups. Blacks and Filipinos were over-represented in our study sample. This fact is consistent with the data of Gray and colleagues who found odds ratios for developing varicella in naval recruits to be 8.8 for Filipinos and 1.9 for blacks (24). This propensity for patients from warmer climates to develop varicella late in life is well known (25), and varicella outbreaks among U.S. Army recruits from Puerto Rico have recently been reported (26). Most of our patients were clustered tightly around 20 years of age; the paucity of subjects over 25 years of age can probably be attributed to the low probability that a seronegative serviceman or woman could consistently avoid exposure. Given the natural history of adult varicella, and the potential for acyclovir to modify it, is treatment worthwhile? The primary benefit of therapy is an earlier return to normal activities. A 7-day course of oral acyclovir at the dose used in our study costs about $100, and for this "price" an average patient will be back to school, work, or military service 1.8 days earlier. In situations where hospitalization for isolation purposes is required (for example, military settings, college infirmaries), 1 or 2 days of hospitalization are also saved. Additional benefits of early treatment include a clinically relevant decrease in the maximum number of skin lesions, fever, and symptoms. Acyclovir therapy of varicella is not without potential problems. The cost is substantial, low but finite risk exists for toxicity, and the dosing schedule is inconvenient. Although the selection of increasing acyclovir-resistant strains of varicella zoster virus is a theoretical concern, a study of 20 varicella zoster virus isolates pre- and post-treatment found no change in sensitivity patterns (27). Results of recent reports have shown resistant isolates in immunosuppressed patients, but this finding has not been reported in immunocompetent patients (28, 29). Presumably, a single, short course of oral acyclovir in normal hosts with varicella would exert little selection pressure in favor of resistance. A varicella vaccine is undergoing extensive evaluation, but even if it becomes widely available will probably not be used frequently in adults because over 90% are already seropositive. Additionally, some vaccine failures and waning immunity will occur in those receiving immunization (30), ensuring that adult varicella will not soon disappear. Oral acyclovir appears to be a reasonable therapeutic option for those immunocompetent patients presenting within 24 hours of rash onset. Intravenous therapy will
1 September 1992 • Annals of Internal Medicine • Volume 117 • Number 5
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19758/ by a University of California San Diego User on 03/05/2017
continue to be the standard therapy for those with complicated disease (pneumonia or encephalitis) and immunocompromised hosts. Larger trials with acyclovir and studies of newer agents will help to better define the therapeutic options for this ubiquitous disease. A portion of this work was presented in abstract form at the 29th Interscience Conference on Antimicrobial Agents and Chemotherapy, Atlanta, Georgia, 1990. The views expressed in this article are those of the authors and do not reflect the official policy or position of the Department of the Navy, Department of Defense, nor the United States Government. Acknowledgments: The authors thank Paul McGuirt, MS, and Anne Clemmer, PhD, of the Burroughs-Wellcome Company for their help with data entry and statistical analysis, Burroughs-Wellcome Company for providing the study drug and placebo, Dr. Ann Arvin for doing the varicella serologies, Dr. Mark Sawyer for his critical review of the manuscript, and Cecilia Cesena and Dena Clarke for their help in preparing the text. Requests for Reprints: Mark R. Wallace, MD, Clinical Investigation Department, San Diego Naval Hospital, San Diego, CA 92134-5000. Current Author Addresses: Drs. Wallace and Oldfield: Department of Internal Medicine, Infectious Disease Division, San Diego Naval Hospital, San Diego, CA 92134-5000. Dr. Bowler: Wausau Medical Center, 2727 Plaza Drive, Wausau, WI 54401. Dr. Murray: 230 Prospect Place, Suite 260, Coronado, CA 92118 Dr. Brodine: Naval Health Research Center, San Diego, CA 921865122. References 1. Guess HA, Broughton DD, Melton JL, Kurland LT. PopulationBased studies of varicella complications. Pediatrics. 1986;78(Suppl): 723-7. 2. Felser JM, Freifeld A. The epidemiology, natural history, and complications of varicella, pp. 223-9. In: Straus SE; moderator. Varicella-zoster virus infections: biology, natural history, treatment and prevention. Ann Intern Med. 1988;108:221-35. 3. Preblud SR. Age-specific risks of varicella complications. Pediatrics. 1981;68:14-7. 4. Krause PR, Straus SE. Zoster and its complications. Hosp Pract. 1990;25:61-76. 5. Dorsky DI, Crumpacker CS. Drugs five years later: acyclovir. Ann Intern Med. 1987;107:859-74. 6. Straus SE. The management of varicella and zoster infections. Infect Dis Clin North Am. 1987;1:367-82. 7. Shulman ST. Acyclovir treatment of disseminated varicella in childhood malignant neoplasms. Am J Dis Child. 1985;139:137-40. 8. Nyerges G, Meszner Z, Gyarmati E, Kerpel-Fronius S. Acyclovir prevents dissemination of varicella in immunocompromised children. J Infect Dis. 1988;157:309-13. 9. Shepp DH, Dandliker PS, Meyers JD. Treatment of varicella-zoster virus infections in severely immunocompromised patients. N Engl J Med. 1986;314:208-12.
10. Balfour HH, Bean B, Laskin OL, Ambinder RF, Meyers JD, Wade J C , et al. Acyclovir halts progression of herpes zoster in immunocompromised patients. N Engl J Med. 1983;308:1448-53. 11. Al-Nakib W, Al-Kandari S, El-Khalik DMA, El-Shirbiny AM. A randomized controlled study of intravenous acyclovir (Zovirax) against placebo in adults with chickenpox. J Infect Dis. 1983; 6(Suppl):49-56. 12. Balfour HH, Kelly JM, Suarez CS, Heussner RC, Englund JA, Crane DD, et al. Acyclovir treatment of varicella in otherwise healthy children. J Pediatr. 1990;116:633-9. 13. Dunkle LM, Arvin AM, Whitley RJ, Rotbert HA, Feder HM, Feldman S, et al. A controlled trial of acyclovir for chickenpox in normal children. N Engl J Med. 1991;325:1539-44. 14. Feder HM. Treatment of adult chickenpox with oral acyclovir. Arch Intern Med. 1990;150:2061-5. 15. Haake DA, Zakowski PC, Haake DL, Bryson YJ. Early treatment with acyclovir for varicella pneumonia in otherwise healthy adults: retrospective controlled study and review. Rev Infect Dis. 1990; 12: 788-97. 16. Asano Y, Itakura N, Kajita Y, Suga S, Yoshikawa T, Yazaki T, et al. Severity of viremia and clinical findings in children with varicella. J Infect Dis. 1990;161:1095-8. 17. Triebwasser JH, Harris RE, Bryant RE, Choades ER. Varicella pneumonia in adults. Medicine. 1967;46:409-21. 18. Amundson DE, Call TB, Mull NH. Review of adult varicella. In Program and Abstract of the 5th Pan-American Congress of the Diseases of the Chest (Abstract No. 124) 1990; San Juan, Puerto Rico. 19. Weber DM, Pellecchia JA. Varicella pneumonia. JAMA. 1965;192: 572-3. 20. Hockberger RS, Rothstein RJ. Varicella pneumonia in adults: a spectrum of disease. Ann Emerg Med. 1986;15:931-4. 21. Sargent EN, Corson MJ, Reilly ED. Varicella pneumonia. California Medicine. 1967;107:141-8. 22. Schlossberg D, Littman M. Varicella pneumonia. Arch Intern Med. 1988;148:1630-2. 23. Pitel PA, McCormick KL, Fitzgerald E, Orson JM. Subclinical hepatic changes in varicella infection. Pediatr. 1980;65:631-3. 24. Gray GC, Palinkas LA, Kelly PW. Increasing incidence of varicella hospitalizations in the United States Army and Navy personnel. Pediatrics. 1990;86:867-73. 25. Christie AB. Chickenpox (varicella). In: Christie AB; ed. Infectious Disease: Epidemiology and Clinical Practice. Edinburgh: ChurchillLivingstone, 1987:353-75. 26. Longfield JN, Winn RE, Gibson RL, Juchav SV, Hoffman PV. Varicella outbreaks in Army recruits from Puerto Rico. Arch Intern Med. 1990;150:970-3. 27. Cole NL, Balfour HH. Varicella-zoster virus does not become more resistant to acyclovir during therapy. J Infect Dis. 1986;153:605-8. 28. Jacobson MA, Berger TG, Fikrig S, Becherer P, Moohr JW, Stanat SC, et al. Acyclovir-resistant varicella zoster virus infection after chronic oral acyclovir therapy in patients with acquired immunodeficiency syndrome (AIDS). Ann Intern Med. 1990;112:187-91. 29. Safrin S, Berger TG, Gilson I, Wolfe PR, Wofsy CB, Mills J, et al. Foscarnet therapy in five patients with AIDS and acyclovir-resistant varicella-zoster virus infection. Ann Intern Med. 1991;115:19-21. 30. Gershon AA, Steinberg SP, LaRussa P, Ferrara A, Hammerschlag M, Gelb L, et al. Immunization of healthy adults with live attenuated varicella vaccine. J Infect Dis. 1988;158:132-7.
1 September 1992 • Annals of Internal Medicine • Volume 117 • Number 5
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19758/ by a University of California San Diego User on 03/05/2017
363
• Objective: To assess the efficacy of oral acyclovir in treating adults with varicella and to describe the natural history of adult varicella. • Design: Double-blind, placebo-controlled randomized trial. • Setting: A naval hospital. • Patients: One hundred forty-eight of 206 consecutive adult active duty Navy and Marine Corps personnel who were hospitalized for isolation and inpatient therapy of varicella and who could be treated within 72 hours of rash onset completed the study. The diagnosis of varicella was confirmed by acute and convalescent serology in 143 of 144 patients with available paired sera. • Intervention: Patients were randomly assigned to receive either acyclovir, 800 mg orally five times per day for 7 days, or an identical placebo. Separate randomization codes were used for patients presenting within 24 hours of rash onset and for those presenting 25 to 72 hours after rash onset. • Measurements: Daily lesion counts, symptom scores, temperature measurements, and laboratory tests were used to monitor the course of the illness. • Results: Early treatment (initiated within 24 hours of rash onset) reduced the total time to (100%) crusting from 7.4 to 5.6 days (P = 0.001) and reduced the maximum number of lesions by 46% (P = 0.04). Duration of fever and severity of symptoms were also reduced by early therapy. Late therapy (25 to 72 hours after rash onset) had no effect on the course of illness. Only four patients had pneumonia, and no encephalitis or mortality was noted. • Conclusions: Early therapy with oral acyclovir decreases the time to cutaneous healing of adult varicella, decreases the duration of fever, and lessens symptoms. Initiation of therapy after the first day of illness is of no value in uncomplicated cases of adult varicella. The low frequency of serious complications of varicella (pneumonia, encephalitis, or death) precluded any evaluation of the possible effect of acyclovir on these outcomes.
Varicella (chickenpox) is typically a benign illness of childhood characterized by a vesicular rash and constitutional symptoms. Although only 1% to 2% of varicella cases occur in adults, the illness is often more severe than in children and is associated with a prolonged course and a 25-fold increased risk for mortality (1-4). Safe and effective therapy might decrease the morbidity of adult varicella and could prove cost effective by returning students and workers to their usual activities more rapidly. Acyclovir, an acyclic guanosine analog, is an antiviral agent with a proven safety record and established activity against the varicella zoster virus (VZV) (5). Blood levels sufficient to inhibit replication of varicella zoster virus are achieved with intravenous acyclovir; levels achieved after oral administration are much lower and may not exceed the in-vitro inhibitory concentration for varicella (6). Several trials have shown the value of intravenous acyclovir in treating varicella zoster virus infections in the immunocompromised host (7-10). Intravenous acyclovir has also been shown to shorten the course of varicella in normal adults (11), but the expense and potential hazards of parenteral therapy can rarely be justified. Trials of oral acyclovir in immunocompetent children with varicella have recently been completed. Fever and duration of illness were reduced when therapy was started within the first 24 hours of illness (12, 13). The results of two small, uncontrolled trials of oral acyclovir in normal adults with varicella have also suggested a benefit (14, 15), but no randomized trial of oral acyclovir in immunocompetent adults has been done. We conducted a double-blind, placebo-controlled trial of oral acyclovir in the treatment of adult varicella. Our goals were to assess the value of therapy, to determine how long after rash onset therapy remained beneficial, and to describe the natural history of adult varicella. Methods Patients Adults (17 years of age or older) who had developed a characteristic varicella rash within the previous 72 hours and who were admitted to the Naval Hospital, San Diego, were eligible for enrollment. Servicemen and women with varicella are routinely hospitalized if they reside on board a ship or in a barracks; this is done both to control infection and to facili-
Annals of Internal Medicine. 1992;117:358-363. From the United States Naval Hospital, San Diego, California. For the current author addresses, see end of text. 358
Drug Generic Name acylovir
Brand Name Zovirax
1 September 1992 • Annals of Internal Medicine • Volume 117 • Number 5
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19758/ by a University of California San Diego User on 03/05/2017
tate patient care. Patients excluded from study participation included those who were pregnant, had received any antiviral therapy during the previous week, had been vaccinated for varicella, were positive for human immunodeficiency virus (HIV), or had received any immunoglobulin products within the previous week. A history of either renal disease or acyclovir intolerance was also a criterion for exclusion. A complete medical history and a physical examination were done at study entry. Written informed consent was obtained from all patients. The study was approved by the hospital's committee for the Protection of Human Subjects. Treatment Regimens Patients were randomly assigned in groups of eight to receive either acyclovir, 800 mg orally five times per day (at 0600, 1000, 1400, 1800, 2200 hours) for 7 days or an identical placebo. All 35 doses were administered by nursing personnel. Two separate randomization codes were used: one for patients admitted within 24 hours of rash onset (early group) and one for patients admitted 25 to 72 hours after rash onset (late group). The patient's best estimate of rash onset was used; prodromal symptoms of headache or fever were not used for grouping patients. Evaluation of Disease Status Clinical evaluations were done at study entry and daily thereafter. The number of lesions within a standard 25x25 cm area (defined as the "pox box") on the chest and back of each patient was determined, and all new lesions were noted. The "pox box" was used to facilitate accurate counting. Because the upper torso often exhibits acneiform lesions, which can be mistaken for late varicella lesions, this area was not used for study evaluation. The percentage of lesions that were maculopapular (erythematous bump without fluid), vesicular (blisterfluid filled or collapsed), crusted, or healed were estimated to the nearest 10 percentage points. Counting and grading of lesions were done daily through day 7, or until all lesions were crusted or healed. Temperature, blood pressure, pulse, and respiratory rate were recorded every 8 hours by nursing personnel. Patients were asked daily about pruritus, anorexia, and lethargy. A quantitative measure of constitutional illness was derived by summing daily the severity scores for anorexia, lethargy, and fever. These complaints were graded as none (0 points), mild (1 point), moderate (2 points), or severe (3 points). Conversion of fever scores to 0, 1, 2, or 3 points corresponded to daily recorded temperatures of less than 37.8 °C, 37.8 to 38.3 °C, 38.4 to 39.4 °C, and 39.4 °C or greater. Laboratory studies on the day of entry assessed the leukocyte count, hematocrit, and platelet count, as well as levels of alanine aminotransferase (ALT), aspartate aminotransferase (AST), total bilirubin, alkaline phosphatase, blood urea nitro-
gen, and serum creatinine. Serum samples were also evaluated for acute varicella serology. Liver function tests were repeated on day 4, and all laboratory tests were repeated on day 7. Convalescent varicella serum was obtained on day 21 (days 14 through 28 were acceptable); all sera was frozen at -70 °C. A chest radiograph was done on day 1. If radiographic results showed an infiltrate, the test was repeated weekly until results were normal. If any laboratory value (except hematocrit) was abnormal on day 7, tests were repeated weekly until normal. Diphenhydramine (50-mg tablets) and acetaminophen (325mg tablets) were available to patients on request to treat pruritus and constitutional symptoms, respectively. Acetaminophen was given every 4 hours to patients whose temperatures were over 38.5 °C. Patients were hospitalized a minimum of 7 days and discharged when all skin lesions were crusted or healed. Serologic Tests Immunoglobulin G (IgG) antibodies to varicella zoster virus were measured by enzyme immunoassay. The wells of microtiter plates (Immulon 2, Dynatech Laboratories, Chantilly, Virginia) were coated with varicella zoster virus antigen or uninfected melanoma cell control and incubated overnight at 4 °C. The plates were washed, and 3% bovine serum albumin in phosphate buffered saline was added to block nonspecific binding. Each test serum was added to duplicate wells and serially diluted from 1:4 to 1:65 536. Biotinylated anti-IgG was added, amplified with an avidin-biotin complex and color change was detected by adding o-phenylene-diamine dihydrochloride and H 2 0 2 (Vectastain ABC Kit, Vector, Burlingame, California). Antibody titers equal to or greater than 1:4 were considered positive. Statistical Analysis The Kaplan-Meier product-limit method was used to estimate all times to events, calculated from the date of study entry to the date of the event. Comparisons of time differences between the treatment groups were done using the log-rank test. Treatment differences were assessed using two-sided tests, and statistical significance was set at P = 0.05. Results were reported separately for patients with rash onset occurring 24 hours or less before study entry and those with onset occurring more than 24 hours before study entry. Ninety-five percent CIs were computed for the relative risk of acyclovir. The relative risk of acyclovir was used to compare the response rate of patients receiving acyclovir with that of patients receiving placebo. Relative risks less than 1.00 indicated that acyclovir performed better than placebo. Differences in medians between treatment groups were compared using the Wilcoxon rank-sum test. Ninety-five percent CIs for the difference are given where appropriate. Percentages of patients in
Table 1. Patient Characteristics at Study Entry Characteristic
Age, y Mean (±SD) Race, n(%) Black White Filipino Other Lesions, n* Mean (±SD) Duration of rash, h Mean (±SD) Temperature at study entry, °C Mean (±SD)
Early Group (< 24 Hours) Acyclovir Placebo in = 38) (n = 38)
Late Group (> 24 to 72 Hours) Placebo Acyclovir (n = 36) (n = 36)
20.26 (2.97)
20.00 (2.51)
21.00 (2.97)
19.67 (2.38)
14 (37) 20 (53) 3(8) 1(3)
13 (34) 17 (45) 5(13) 3(8)
10 (28) 17 (47) 7(19) 2(6)
13 (36) 13 (36) 8(22) 2(6)
193 (170)
199 (220)
248 (226)
198 (187)
17.2 (6.2)
18.8 (5.5)
44.3 (11.64)
45.4 (11.34)
37.8 (1.6)
37.4 (1.5)
37.7 (1.6)
37.6 (1.5)
* Within the "pox box" area; chest and back. See text for description. 1 September 1992 • Annals of Internal Medicine • Volume 117 • Number 5
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19758/ by a University of California San Diego User on 03/05/2017
359
Table 2 . Cutaneous Effects of Varicella Variable
Time to maximum number of skin lesions, d Time of new lesion formation, d Time to onset of cutaneous healing, d Time to 100% crusting, d Maximum number of lesions, n*
Early Group (< 24 Hours) Acyclovir Placebo P Value (n = 38) (n = 38) 1.5 2.7 2.6 5.6 268
2.1 3.3 3.3 7.4 500
0.002 0.03 < 0.001 0.001 0.04
Late Group (25 to 72 Hours) Acyclovir Placebo P Value {n = 36) (n = 36) 1.3 3.0 2.4 7.0 233
1.2 2.3 2.3 6.8 158
> 0.2 0.03 > 0.2 > 0.2 0.03
* Wilcoxon rank-sum test. All other comparisons by log-rank test with median values from Kaplan-Meier product-limit plots.
each treatment group were compared using the Fisher exact test. Results Study Patients All 206 inpatients with varicella who were admitted between 15 February 1989 and 15 March 1990 were evaluated for enrollment. Fifty-three patients were excluded: 44 had rash onset more than 72 hours before hospitalization, 8 refused to participate, and 1 was pregnant. Of the remaining patients, four were withdrawn after 1 to 2 days because of incorrect diagnosis (all had vaccinia secondary to recent smallpox vaccination), and one was withdrawn after developing hives on day 5 of therapy. These 5 patients were excluded from all analyses, leaving 148 evaluable patients. Seventy-six patients were admitted within 24 hours of rash onset (early group), and 72 patients were admitted 25 to 72 hours after rash onset (late group). Only three patients were women. Most of our patients were Navy or Marine recruits undergoing basic training (women recruits receive basic training elsewhere). Characteristics of the evaluable patients are shown in Table 1. The mean age of study patients was 20 years (range, 17 to 33 years). The placebo and treatment groups and the early and late subgroups were similar with respect to initial lesion counts, duration of rash before study entry, and temperature at the time of hospitalization.
Cutaneous Effects Five parameters of cutaneous disease progression were evaluated. By all these criteria, early acyclovir treatment was beneficial, whereas delayed therapy (25 to 72 hours after rash onset) had no positive effect. Early treatment shortened the time to maximum number of lesions from 2.1 to 1.5 days (P = 0.002; relative risk, 0.60; 95% CI, 0.37 to 0.97) and reduced the days of new lesion formation from 3.3 to 2.7 days (P = 0.03; relative risk, 0.70; 95% CI, 0.44 to 1.11). In addition, early acyclovir therapy sped the onset of cutaneous healing (defined as the first day the total number of lesions decreased by 5% and continued to decline thereafter) from 3.3 to 2.6 days (P < 0.001; relative risk, 0.54; 95% CI, 0.33 to 0.88), and produced full crusting in 5.6 compared with 7.4 days (P = 0.001; relative risk, 0.56; 95% CI, 0.34 to 0.93). The maximum number of lesions was also reduced by 46% in the early group (P = 0.04). When the late group was considered, acyclovir had no beneficial effect on any parameter (Table 2). The proportion of patients requiring more than 7 days to achieve 100% crusting was significantly reduced in those patients who were treated within 24 hours. Of 38 patients receiving acyclovir, 12 (38%) took more than 7 days to crust, compared with 22 of 38 patients (58%) receiving placebo (P = 0.02 by chi square). No such benefit was observed in the late group. Effect on Fever For the group enrolled within 24 hours of rash onset, acyclovir treatment reduced the duration of fever from 2.74 to 2.24 days (P = 0.02), (Figure 1). No treatment effect was observed in the late group (2.1 compared with 2.04 days, P = 0.65). Six of 38 patients (16%) treated with acyclovir within 24 hours were febrile for more than 72 hours, whereas 17 of 38 early placebo recipients (45%) were febrile for more than 72 hours (P = 0.02, chi square). Effect on Symptoms and Symptomatic Therapy
Figure 1. Proportion of patients with fever in the early treatment group, acyclovir compared with placebo.
360
Pruritus tended to be less severe and to require less diphenhydramine among patients treated with acyclovir in the early group. This trend was significant only on day 5 (P = 0.025, Fisher exact test); late treatment had no effect. In addition, acetaminophen consumption was consistently less frequent in the early acyclovir treat-
1 September 1992 • Annals of Internal Medicine • Volume 117 • Number 5
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19758/ by a University of California San Diego User on 03/05/2017
ment group. This difference was statistically significant only on days 4 and 5 (P - 0.04 and 0.03, respectively; Fisher exact test). By day 5, only 2.6% of the early acyclovir group was using acetaminophen compared with 18% of the early placebo group. The mean constitutional illness scores peaked on the second day and decreased thereafter for all patients. Although the acyclovir-treated patients had higher mean admission scores, their symptoms diminished more rapidly than did those of the placebo treated groups. This effect was especially pronounced in the early group (Figure 2).
Toxicity One patient who was receiving acyclovir developed urticaria on day 4 and was removed from the study. Another patient who was receiving acyclovir complained of transient nausea but did not require dose modification. Renal abnormalities were not observed. All patients had normal blood urea nitrogen and creatinine levels on both days 1 and 7, except one placebo patient who had a mildly elevated blood urea nitrogen at the time of hospitalization; this abnormality resolved by day 7. Serology
Complications Pulmonary infiltrates were seen on the admission chest radiographs of four patients from the late group, two patients each in the acyclovir and placebo groups. All infiltrates resolved during follow-up. No patient hospitalized with a normal chest radiograph developed pulmonary complaints of sufficient severity to warrant a repeat procedure. Late complications of varicella were infrequent. Two placebo recipients developed secondary skin infections severe enough to require antibiotic therapy. No meningeal or encephalitic manifestations were noted in any patient. Asymptomatic liver function abnormalities were common, and no difference in the frequency, severity, or duration of asymptomatic hepatitis was seen between patients in the acyclovir and placebo groups. Fifty percent of all patients had elevated ALT or AST levels on day 1, 4, or 7. Elevations were generally highest on day 4 or 7 and resolved in 1 to 5 weeks. One patient's ALT and AST levels peaked at 14 times the upper limit of normal on day 7 and slowly decreased to normal over 5 months; an extensive investigation disclosed no other etiology for his prolonged asymptomatic hepatitis. Our patients' hepatic abnormalities were similar to other viral hepatitides in that the ALT level usually exceeded that of the AST. In only five patients did either the ALT or AST level exceed 5 times the upper limit of normal. Slight elevations of total bilirubin were seen in four patients, three of whom had transaminase elevations; none was jaundiced. Alkaline phosphatase levels were mildly increased (1.0 to 1.5 times the upper limit of normal) in three patients. The frequency, duration, and severity of the varicella-associated anemia, thrombocytopenia, and leukopenia were unaffected by acyclovir therapy. Mild thrombocytopenia (platelet count 100 to 150 x 109/L) occurred in 10 of 148 patients (6.8%). Thrombocytopenia resolved before hospital discharge in all cases and was unassociated with bleeding or petechiae. Forty-one patients (27.7%) had a leukocyte count of less than 4.8 x 109/L at the time of hospitalization; the lowest count was 2.5 x 10 9 /L. All cases of leukopenia resolved within 1 week without sequelae. Mild anemia (hematocrit, 0.35 to 0.41 for men) occurred in 30 of 148 patients (20.3%); in only 4 of these patients was the hematocrit less than 0.38.
Acute and convalescent sera were available from 144 of the 148 patients who completed the study. Seroconversion, defined as a two-dilution increase in titer, was documented in 143 of these patients. One patient who was hospitalized with a 3-day history of rash had a high but unchanging titer (1:1024). There was no significant difference in the mean convalescent titers of the acyclovir and placebo groups. Discussion Adult varicella, although uncommon, remains a serious problem and accounts for one fourth of all varicellarelated fatalities in the United States (1, 2). Oral acyclovir therapy reduced the duration of varicella in immunocompetent children when therapy was begun within 24 hours of rash onset (12, 13), but its potential benefit in adults has been unproven. We showed that early acyclovir therapy significantly affected the clinical course of varicella. Time to 100% crusting was reduced by nearly 2 days, and days of fever were reduced by one half day. The reduction in time to 100% crusting was clinically significant in that these sailors and marines were able to return to full duty earlier. All four secondary parameters of cutaneous healing (maximum number of skin lesions, time to onset of cutaneous healing, days of new lesion formation, and days to maximum number of lesions) were also significantly improved by early acyclovir therapy. Significant treatment differences were observed in the
Figure 2. Daily constitutional illness scores of patients treated within 24 hours.
1 September 1992 • Annals of Internal Medicine • Volume 117 • Number 5
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19758/ by a University of California San Diego User on 03/05/2017
361
resolution of constitutional symptoms and signs associated with varicella in the acyclovir-treated patients in the early group. No benefits in reduction of fever, cutaneous involvement, symptoms, or duration of illness were observed in the late group. The results of our study are consistent with those of the acyclovir treatment studies recently completed in normal children (12, 13), which found a similar benefit of acyclovir when started within 24 hours of rash onset, and show that later initiation of therapy is not useful. One explanation for this lack of treatment utility after the first 24 hours of rash onset may be the pattern of viremia in varicella, which peaks just before onset of rash in immunocompetent persons (16). Acyclovir treatment later than 24 hours after onset may be superfluous because viral replication is already being rapidly controlled by the intact host immune system. In immunosuppressed patients with varicella zoster virus infections, however, later initiation of therapy (more than 72 hours after symptom onset) may be of value (10). These patients have presumably sustained viral replication and dissemination related to their immune deficits, allowing acyclovir therapy to yield a treatment benefit much later in the course of their illness. We also observed that patients in the placebo group who presented on day 1 had a more protracted illness than did those who presented on days 2 and 3. Perhaps the patients with more serious disease seek evaluation earlier and thus have more to gain from early intervention. Four patients (2.7%) had radiographically confirmed varicella pneumonia at the time of hospitalization. No patient developed signs or symptoms of pneumonia more than 3 days after rash onset. This result is consistent with those of a previous review, which found 70% of cases of varicella pneumonia occurred within the first 3 days of illness (17). The relatively low incidence of varicella pneumonia in our patients is consistent with the 4.5% found in a recent 485-patient retrospective review of varicella (18). A 1967 study of U.S. Navy personnel found a 16.3% incidence of radiographically diagnosed pneumonia (19); other series have found the incidence of varicella pneumonia in adults with varicella to be as great as 50% (20). In our study, all four patients with pneumonia had a dry cough, but none was dyspneic or hypoxic. Their clinical course was benign, and the radiographic abnormalities resolved within 2 to 3 weeks without sequelae. Although most patients with varicella pneumonia are asymptomatic or have only a dry cough and recover uneventfully, severe cases can occur. Fatal varicella pneumonia has been reported in both immunocompetent (21) and immunocompromised hosts (2) as well as in pregnant women (20). Pneumonia represents the most frequent severe complication of adult varicella. Treatment with acyclovir has not been prospectively studied; however, limited information suggests a benefit (22). We recommend intravenous acyclovir for patients with progressive or symptomatic pneumonia; asymptomatic patients (or those with only a mild cough) should receive oral acyclovir if they present within the first day of their illness. Half of our patients had asymptomatic hepatitis as shown by elevation of either their ALT or AST levels. 362
This is consistent with previous natural history data that showed a 77% incidence of "subclinical hepatitis" (23). Routine testing for hepatic dysfunction in the immunocompetent host with varicella is probably not indicated. The discovery of mild, asymptomatic hepatitis is not unexpected and deserves no particular follow-up or therapy. Acyclovir-related adverse events were rare. One patient developed hives and was dropped from the study. Although rashes have previously been reported among patients receiving acyclovir (5), they have been infrequent. Renal toxicity was not observed, and gastrointestinal complaints were equally infrequent in both the treated and placebo groups. Blacks and Filipinos were over-represented in our study sample. This fact is consistent with the data of Gray and colleagues who found odds ratios for developing varicella in naval recruits to be 8.8 for Filipinos and 1.9 for blacks (24). This propensity for patients from warmer climates to develop varicella late in life is well known (25), and varicella outbreaks among U.S. Army recruits from Puerto Rico have recently been reported (26). Most of our patients were clustered tightly around 20 years of age; the paucity of subjects over 25 years of age can probably be attributed to the low probability that a seronegative serviceman or woman could consistently avoid exposure. Given the natural history of adult varicella, and the potential for acyclovir to modify it, is treatment worthwhile? The primary benefit of therapy is an earlier return to normal activities. A 7-day course of oral acyclovir at the dose used in our study costs about $100, and for this "price" an average patient will be back to school, work, or military service 1.8 days earlier. In situations where hospitalization for isolation purposes is required (for example, military settings, college infirmaries), 1 or 2 days of hospitalization are also saved. Additional benefits of early treatment include a clinically relevant decrease in the maximum number of skin lesions, fever, and symptoms. Acyclovir therapy of varicella is not without potential problems. The cost is substantial, low but finite risk exists for toxicity, and the dosing schedule is inconvenient. Although the selection of increasing acyclovir-resistant strains of varicella zoster virus is a theoretical concern, a study of 20 varicella zoster virus isolates pre- and post-treatment found no change in sensitivity patterns (27). Results of recent reports have shown resistant isolates in immunosuppressed patients, but this finding has not been reported in immunocompetent patients (28, 29). Presumably, a single, short course of oral acyclovir in normal hosts with varicella would exert little selection pressure in favor of resistance. A varicella vaccine is undergoing extensive evaluation, but even if it becomes widely available will probably not be used frequently in adults because over 90% are already seropositive. Additionally, some vaccine failures and waning immunity will occur in those receiving immunization (30), ensuring that adult varicella will not soon disappear. Oral acyclovir appears to be a reasonable therapeutic option for those immunocompetent patients presenting within 24 hours of rash onset. Intravenous therapy will
1 September 1992 • Annals of Internal Medicine • Volume 117 • Number 5
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19758/ by a University of California San Diego User on 03/05/2017
continue to be the standard therapy for those with complicated disease (pneumonia or encephalitis) and immunocompromised hosts. Larger trials with acyclovir and studies of newer agents will help to better define the therapeutic options for this ubiquitous disease. A portion of this work was presented in abstract form at the 29th Interscience Conference on Antimicrobial Agents and Chemotherapy, Atlanta, Georgia, 1990. The views expressed in this article are those of the authors and do not reflect the official policy or position of the Department of the Navy, Department of Defense, nor the United States Government. Acknowledgments: The authors thank Paul McGuirt, MS, and Anne Clemmer, PhD, of the Burroughs-Wellcome Company for their help with data entry and statistical analysis, Burroughs-Wellcome Company for providing the study drug and placebo, Dr. Ann Arvin for doing the varicella serologies, Dr. Mark Sawyer for his critical review of the manuscript, and Cecilia Cesena and Dena Clarke for their help in preparing the text. Requests for Reprints: Mark R. Wallace, MD, Clinical Investigation Department, San Diego Naval Hospital, San Diego, CA 92134-5000. Current Author Addresses: Drs. Wallace and Oldfield: Department of Internal Medicine, Infectious Disease Division, San Diego Naval Hospital, San Diego, CA 92134-5000. Dr. Bowler: Wausau Medical Center, 2727 Plaza Drive, Wausau, WI 54401. Dr. Murray: 230 Prospect Place, Suite 260, Coronado, CA 92118 Dr. Brodine: Naval Health Research Center, San Diego, CA 921865122. References 1. Guess HA, Broughton DD, Melton JL, Kurland LT. PopulationBased studies of varicella complications. Pediatrics. 1986;78(Suppl): 723-7. 2. Felser JM, Freifeld A. The epidemiology, natural history, and complications of varicella, pp. 223-9. In: Straus SE; moderator. Varicella-zoster virus infections: biology, natural history, treatment and prevention. Ann Intern Med. 1988;108:221-35. 3. Preblud SR. Age-specific risks of varicella complications. Pediatrics. 1981;68:14-7. 4. Krause PR, Straus SE. Zoster and its complications. Hosp Pract. 1990;25:61-76. 5. Dorsky DI, Crumpacker CS. Drugs five years later: acyclovir. Ann Intern Med. 1987;107:859-74. 6. Straus SE. The management of varicella and zoster infections. Infect Dis Clin North Am. 1987;1:367-82. 7. Shulman ST. Acyclovir treatment of disseminated varicella in childhood malignant neoplasms. Am J Dis Child. 1985;139:137-40. 8. Nyerges G, Meszner Z, Gyarmati E, Kerpel-Fronius S. Acyclovir prevents dissemination of varicella in immunocompromised children. J Infect Dis. 1988;157:309-13. 9. Shepp DH, Dandliker PS, Meyers JD. Treatment of varicella-zoster virus infections in severely immunocompromised patients. N Engl J Med. 1986;314:208-12.
10. Balfour HH, Bean B, Laskin OL, Ambinder RF, Meyers JD, Wade J C , et al. Acyclovir halts progression of herpes zoster in immunocompromised patients. N Engl J Med. 1983;308:1448-53. 11. Al-Nakib W, Al-Kandari S, El-Khalik DMA, El-Shirbiny AM. A randomized controlled study of intravenous acyclovir (Zovirax) against placebo in adults with chickenpox. J Infect Dis. 1983; 6(Suppl):49-56. 12. Balfour HH, Kelly JM, Suarez CS, Heussner RC, Englund JA, Crane DD, et al. Acyclovir treatment of varicella in otherwise healthy children. J Pediatr. 1990;116:633-9. 13. Dunkle LM, Arvin AM, Whitley RJ, Rotbert HA, Feder HM, Feldman S, et al. A controlled trial of acyclovir for chickenpox in normal children. N Engl J Med. 1991;325:1539-44. 14. Feder HM. Treatment of adult chickenpox with oral acyclovir. Arch Intern Med. 1990;150:2061-5. 15. Haake DA, Zakowski PC, Haake DL, Bryson YJ. Early treatment with acyclovir for varicella pneumonia in otherwise healthy adults: retrospective controlled study and review. Rev Infect Dis. 1990; 12: 788-97. 16. Asano Y, Itakura N, Kajita Y, Suga S, Yoshikawa T, Yazaki T, et al. Severity of viremia and clinical findings in children with varicella. J Infect Dis. 1990;161:1095-8. 17. Triebwasser JH, Harris RE, Bryant RE, Choades ER. Varicella pneumonia in adults. Medicine. 1967;46:409-21. 18. Amundson DE, Call TB, Mull NH. Review of adult varicella. In Program and Abstract of the 5th Pan-American Congress of the Diseases of the Chest (Abstract No. 124) 1990; San Juan, Puerto Rico. 19. Weber DM, Pellecchia JA. Varicella pneumonia. JAMA. 1965;192: 572-3. 20. Hockberger RS, Rothstein RJ. Varicella pneumonia in adults: a spectrum of disease. Ann Emerg Med. 1986;15:931-4. 21. Sargent EN, Corson MJ, Reilly ED. Varicella pneumonia. California Medicine. 1967;107:141-8. 22. Schlossberg D, Littman M. Varicella pneumonia. Arch Intern Med. 1988;148:1630-2. 23. Pitel PA, McCormick KL, Fitzgerald E, Orson JM. Subclinical hepatic changes in varicella infection. Pediatr. 1980;65:631-3. 24. Gray GC, Palinkas LA, Kelly PW. Increasing incidence of varicella hospitalizations in the United States Army and Navy personnel. Pediatrics. 1990;86:867-73. 25. Christie AB. Chickenpox (varicella). In: Christie AB; ed. Infectious Disease: Epidemiology and Clinical Practice. Edinburgh: ChurchillLivingstone, 1987:353-75. 26. Longfield JN, Winn RE, Gibson RL, Juchav SV, Hoffman PV. Varicella outbreaks in Army recruits from Puerto Rico. Arch Intern Med. 1990;150:970-3. 27. Cole NL, Balfour HH. Varicella-zoster virus does not become more resistant to acyclovir during therapy. J Infect Dis. 1986;153:605-8. 28. Jacobson MA, Berger TG, Fikrig S, Becherer P, Moohr JW, Stanat SC, et al. Acyclovir-resistant varicella zoster virus infection after chronic oral acyclovir therapy in patients with acquired immunodeficiency syndrome (AIDS). Ann Intern Med. 1990;112:187-91. 29. Safrin S, Berger TG, Gilson I, Wolfe PR, Wofsy CB, Mills J, et al. Foscarnet therapy in five patients with AIDS and acyclovir-resistant varicella-zoster virus infection. Ann Intern Med. 1991;115:19-21. 30. Gershon AA, Steinberg SP, LaRussa P, Ferrara A, Hammerschlag M, Gelb L, et al. Immunization of healthy adults with live attenuated varicella vaccine. J Infect Dis. 1988;158:132-7.
1 September 1992 • Annals of Internal Medicine • Volume 117 • Number 5
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19758/ by a University of California San Diego User on 03/05/2017
363
