Oral antiviral therapy for prevention of genital herpes outbreaks in immunocompetent and nonpregnant patients

  • Review
  • Intervention

Authors


Abstract

Background

Genital herpes is caused by herpes simplex virus 1 (HSV-1) or 2 (HSV-2). Some infected people experience outbreaks of genital herpes, typically, characterized by vesicular and erosive localized painful genital lesions.

Objectives

To compare the effectiveness and safety of three oral antiviral drugs (acyclovir, famciclovir and valacyclovir) prescribed to suppress genital herpes outbreaks in non-pregnant patients.

Search methods

We searched the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE, EMBASE, the search portal of the World Health Organization International Clinical Trials Registry Platform and pharmaceutical company databases up to February 2014. We also searched US Food and Drug Administration databases and proceedings of seven congresses to a maximum of 10 years. We contacted trial authors and pharmaceutical companies.

Selection criteria

We selected parallel-group and cross-over randomized controlled trials including patients with recurrent genital herpes caused by HSV, whatever the type (HSV-1, HSV-2, or undetermined), with at least four recurrences per year (trials concerning human immunodeficiency virus (HIV)-positive patients or pregnant women were not eligible) and comparing suppressive oral antiviral treatment with oral acyclovir, famciclovir, and valacyclovir versus placebo or another suppressive oral antiviral treatment.

Data collection and analysis

Two review authors independently selected eligible trials and extracted data. The Risk of bias tool was used to assess risk of bias. Treatment effect was measured by the risk ratio (RR) of having at least one genital herpes recurrence. Pooled RRs were derived by conventional pairwise meta-analyses. A network meta-analysis allowed for estimation of all possible two-by-two comparisons between antiviral drugs.

Main results

A total of 26 trials (among which six had a cross-over design) were included. Among the 6950 randomly assigned participants, 54% (range 0 to100%) were female, mean age was 35 years (range 26 to 45.1), and the mean number of recurrences per year was 11 (range 6.3 to 17.8). Duration of treatment was two to 12 months. Risk of bias was considered high for half of the studies and unclear for the other half. A total of 14 trials compared acyclovir versus placebo, four trials compared valacyclovir versus placebo and 2 trials compared valacyclovir versus no treatment. Three trials compared famciclovir versus placebo. Two trials compared valacyclovir versus famciclovir and one trial compared acyclovir versus valacyclovir versus placebo.

We analyzed data from 22 trials for the outcome: risk of having at least one clinical recurrence. We could not obtain the outcome data for four trials. In placebo-controlled trials, there was a low quality evidence that the risk of having at least one clinical recurrence was reduced with acyclovir (nine parallel-group trials, n = 2049; pooled RR 0.48, 95% confidence interval (CI) 0.39 to 0.58), valacyclovir (four trials, n = 1788; pooled RR 0.41, 95% CI 0.24 to 0.69), or famciclovir (two trials, n = 732; pooled RR 0.57, 95% CI 0.50 to 0.64). The six cross-over trials showed larger treatment effects on average than the parallel-group trials. We found evidence of a small-study effect for acyclovir placebo-controlled trials (adjusted pooled RR 0.61, 95% CI 0.49 to 0.75). In analyzing parallel-group trials by daily dose, no clear evidence was found of a dose-response relationship for any drug. In head-to-head trials, the risk of having at least one recurrence was increased with valacyclovir rather than acyclovir (one trial, n = 1345; RR 1.16, 95% CI 1.01 to 1.34) and was not significantly different from that seen with famciclovir as compared with valacyclovir (one trial, n = 320; RR 1.18, 95% CI 0.86 to 1.63).

We included 16 parallel-arm trials in a network meta-analysis and we were unable to determine which of the drugs was most effective in reducing the risk of at least one clinical recurrence (after adjustment for small-study effects, pooled RR 0.83, 95% CI 0.61 to 1.11 for valacyclovir vs acyclovir; pooled RR 1.04, 95% CI, 0.71 to 1.49 for famciclovir vs acyclovir; and pooled RR 1.26, 95% CI 0.89 to 1.75 for famciclovir vs valacyclovir). Safety data were sought but were reported as total numbers of adverse events.

Authors' conclusions

Owing to risk of bias and inconsistency, there is low quality evidence that suppressive antiviral therapy with acyclovir, valacyclovir or famciclovir in pacients experiencing at least four recurrences of genital herpes per year decreases the number of pacients with at least one recurrence as compared with placebo. Network meta-analysis of the few direct comparisons and the indirect comparisons did not show superiority of one drug over another.

Résumé scientifique

Traitement antiviral oral pour la prévention des poussées d'herpès génital chez les patient(e)s immunocompétent(e)s en l'absence de grossesse

Contexte

L'herpès génital est causé par le virus de l'herpès simplex de type 1 (VHS-1) ou 2 (VHS-2). Certaines personnes infectées connaissent des flambées d'herpès génital, généralement caractérisées par des lésions vésiculaires et érosives génitales localisées et douloureuses.

Objectifs

Comparer l'efficacité et l'innocuité de trois médicaments antiviraux oraux (aciclovir, famciclovir et valaciclovir) prescrits pour supprimer les poussées d'herpès génital chez les patient(e)s en l'absence de grossesse.

Stratégie de recherche documentaire

Nous avons effectué des recherches dans le registre Cochrane des essais contrôlés (CENTRAL), MEDLINE, EMBASE, le portail de recherche du Système d'enregistrement international des essais cliniques (ICTRP) de l'Organisation mondiale de la Santé (OMS) et les bases de données des laboratoires pharmaceutiques jusqu'à février 2014. Nous avons également cherché dans les bases de données de la US Food and Drug Administration et dans les actes de sept congrès, en remontant à 10 ans en arrière au maximum. Nous avons également contacté des auteurs d'essais et des laboratoires pharmaceutiques.

Critères de sélection

Nous avons sélectionné des essais contrôlés randomisés en groupes parallèles et croisés, incluant des patients atteints d'herpès génital récidivant causé par le VHS quel qu'en soit le type (VHS-1, VHS-2 ou indéterminé), ayant au moins quatre récidives par an (nous n'avons pas retenu les essais menés sur des patients positifs pour le virus de l'immunodéficience humaine (VIH) ni sur des femmes enceintes) et comparant le traitement antiviral suppresseur par l'aciclovir, le famciclovir ou le valaciclovir par voie orale avec un placebo ou un autre traitement antiviral suppresseur oral.

Recueil et analyse des données

Deux auteurs de la revue ont sélectionné les essais et extrait les données de manière indépendante. Le risque de biais a été évalué à l'aide de l'outil correspondant. L'effet du traitement a été mesuré par le risque relatif (RR) d'avoir au moins une récidive d'herpès génital. Des RR groupés ont été obtenus par des méta-analyses par paires classiques. Une méta-analyse en réseau a permis d'estimer toutes les comparaisons deux par deux possibles entre les médicaments antiviraux.

Résultats principaux

Au total, 26 essais (dont six de conception croisée) ont été inclus. Parmi les 6 950 participants randomisés, 54 % (extrêmes de 0 et 100 %) étaient des femmes, l'âge moyen était de 35 ans (extrêmes de 26 et 45,1), et le nombre moyen de récidives par an était de 11 (fourchette de 6,3 à 17,8). La durée du traitement était de 2 à 12 mois. Le risque de biais a été jugé élevé pour la moitié des études et incertain pour l'autre moitié. Au total, 14 essais ont comparé l'aciclovir à un placebo, 4 essais ont comparé le valaciclovir à un placebo et 2 essais le valaciclovir à l'absence de traitement. Trois essais ont comparé le famciclovir à un placebo. Deux essais ont comparé le valaciclovir au famciclovir et un essai l'aciclovir au valaciclovir et à un placebo.

Nous avons analysé les données de 22 essais en définissant comme critère d'évaluation le risque d'avoir au moins une récidive clinique. Nous n'avons pas pu obtenir de données sur ce paramètre dans quatre essais. Dans les essais contrôlés par placebo, il existait des preuves de faible qualité d'une réduction du risque d'avoir au moins une récidive clinique avec l'aciclovir (neuf essais en groupes parallèles, n = 2 049 ; RR groupé 0,48, intervalle de confiance (IC) à 95 % de 0,39 à 0.58), le valaciclovir (quatre essais, n = 1 788 ; RR groupé 0,41, IC à 95 % de 0,24 à 0,69), ou le famciclovir (deux essais, n = 732 ; RR groupé 0,57, IC à 95 % de 0,50 au 0,64). Les six essais croisés ont montré, en moyenne, des effets plus importants du traitement que les essais en groupes parallèles. Nous avons trouvé des preuves d'un effet « petites études » pour les essais contrôlés par placebo de l'aciclovir (RR groupé ajusté 0,61, IC à 95 % de 0,49 à 0,75). En analysant les essais en groupes parallèles par dose quotidienne, nous n'avons trouvé aucune preuve claire d'une relation dose-réponse pour l'un des médicaments. Dans les essais de comparaison directe, le risque d'avoir au moins une récidive était augmenté avec le valaciclovir plus qu'avec l'aciclovir (un essai, n = 1 345 ; RR 1,16, IC à 95 % de 1,01 à 1,34) et n'a pas été significativement différent de celui observé avec le famciclovir par rapport au valaciclovir (un essai, n = 320 ; RR 1,18, IC à 95 % de 0,86 à 1,63).

Nous avons inclus 16 essais en groupes parallèles dans une méta-analyse en réseau, sans pouvoir déterminer lequel des médicaments était le plus efficace pour réduire le risque d'au moins une récidive clinique (après prise en compte de l'effet « petites études », RR groupé 0,83, IC à 95 % de 0,61 à 1,11 pour la comparaison valaciclovir-aciclovir ; RR groupé 1,04, IC à 95 % de 0,71 à 1,49 pour famciclovir-aciclovir, et RR groupé 1,26, IC à 95 % de 0,89 à 1,75 pour famciclovir-valaciclovir). Nous avons cherché des données sur l'innocuité mais elles étaient rapportées sous la forme du nombre total d'événements indésirables.

Conclusions des auteurs

Il existe des preuves de faible qualité (en raison du risque de biais et des incohérences) que le traitement antiviral suppresseur par l'aciclovir, le valaciclovir ou le famciclovir chez les patients souffrant d'au moins quatre récidives d'herpès génital par an diminue le nombre de patients ayant au moins une récidive par rapport au placebo. La méta-analyse en réseau des quelques comparaisons directes et des comparaisons indirectes n'a pas montré de supériorité d'un médicament par rapport à un autre.

Notes de traduction

Traduction réalisée par le Centre Cochrane Français

Plain language summary

(Oral antiviral treatment to prevent genital herpes outbreaks in immunocompetent and nonpregnant patients)

Review question

Are oral antiviral drugs (acyclovir, famciclovir, and valacyclovir) effective compared with placebo? And is one of these three drugs superior to the others in suppressing genital herpes outbreaks in patients experiencing four or more recurrences per year? Effectiveness in this review was evaluated by determining the risk of experiencing at least one recurrence during the treatment period in each group.

Background

Genital herpes is a sexually transmitted disease (STD) related to herpes simplex virus type 1 (HSV-1) or 2 (HSV-2). In some people infected with this virus, painful mucocutaneous vesicles develop in a small zone of the genital area and evolve into erosions and crusts. The repetition of this event is called recurrence. Each recurrence lasts five to 10 days.Treatment options in patients experiencing recurrences of genital herpes include no treatment, symptomatic treatment, episodic antiviral treatment for a few days each time a recurrence occurs, and suppressive daily continuous treatment.

Study characteristics

A total of 26 trials including 6950 patients were included in this review. Fifty-four percent of these patients were female, mean age was 35 years, and mean number of recurrences per year before entry into the trials was 11. Duration of treatment in trials ranged from two to 12 months. A total of 14 trials compared acyclovir versus placebo. Four trials compared valacyclovir versus placebo and two trials compared valacyclovir versus no treatment. Three trials compared famciclovir versus placebo. Two trials compared valacyclovir versus famciclovir, and one trial compared acyclovir versus valacyclovir versus placebo. Among the 26 included trials, 22 declared pharmaceutical company funding. The last search for studies was carried out in February 2014.

Key results

Suppressive antiviral therapy with acyclovir, valacyclovir, or famciclovir in patients experiencing at least four recurrences per year decreases the number of patients having at least one recurrence compared with placebo. There is no evidence that suggests that any of these drugs is superior to the others.

Quality of evidence

Althought the three antiviral drugs showed better results compared with placebo, we are uncertain as to how much a difference there are likely to make, because of issues with the conduct and reporting of studies, and inconsistency of their results. The quality of the evidence is low and we think that the size of the effects is likely to change with more research. Because few studies compared the three drugs against one other, we are moderately confident in the fact that there is no difference between the three drugs in terms of effectiveness.

Résumé simplifié

Traitement antiviral oral pour prévenir les poussées d'herpès génital chez les patient(e)s immunocompétent(e)s en l'absence de grossesse

Question de la revue

Les médicaments antiviraux oraux (aciclovir, famciclovir et valaciclovir) sont-ils efficaces par rapport au placebo ? L'un de ces trois médicaments est-il supérieur aux deux autres pour supprimer les poussées d'herpès génital chez les patients souffrant de quatre récidives ou plus par an ? Dans cette revue, l'efficacité a été évaluée en déterminant le risque de connaître au moins une récidive au cours de la période de traitement dans chaque groupe.

Contexte

L'herpès génital est une maladie sexuellement transmissible (MST) causée par le virus de l'herpès simplex de type 1 (HSV-1) ou 2 (HSV-2). Chez certaines personnes infectées par ce virus, des vésicules cutanéo-muqueuses douloureuses se développent dans une petite zone de la région génitale et évoluent en érosions et croûtes. La répétition de cet événement est appelée récidive. Chaque récidive dure de 5 à 10 jours. Les options de traitement en cas de récidive de l'herpès génital comprennent l'abstention thérapeutique, un traitement symptomatique, un traitement antiviral épisodique de quelques jours à chaque récidive et un traitement suppresseur continu quotidien.

Caractéristiques des études

Un total de 26 essais incluant 6 950 patients ont été inclus dans cette revue. Cinquante-quatre pour cent de ces patients étaient des femmes, l'âge moyen était de 35 ans, et le nombre moyen de récidives par an avant l'inclusion dans les essais était de 11. La durée du traitement était de deux à douze mois dans les essais. Au total, 14 essais ont comparé l'aciclovir à un placebo. Quatre essais ont comparé le valaciclovir à un placebo et deux essais le valaciclovir à l'absence de traitement. Trois essais ont comparé le famciclovir à un placebo. Deux essais ont comparé le valaciclovir au famciclovir et un essai l'aciclovir au valaciclovir et à un placebo. Parmi les 26 essais inclus, 22 déclaraient un financement par un laboratoire pharmaceutique. La dernière recherche d'études a été réalisée en février 2014.

Principaux résultats

Le traitement antiviral suppresseur par l'aciclovir, le valaciclovir ou le famciclovir chez les patients souffrant d'au moins quatre récidives par an diminue le nombre de patients ayant au moins une récidive par rapport au placebo. Il n'y a aucune preuve suggérant que l'un de ces médicaments soit supérieur aux autres.

Qualité des preuves

Bien que les trois antiviraux aient donné de meilleurs résultats que le placebo, nous sommes dans l'incertitude quant à l'ampleur de la différence qui peut leur être attribuée, en raison de problèmes de conduite et de compte-rendu des études et d'incohérence de leurs résultats. La qualité des preuves est mauvaise et nous pensons que la taille des effets pourrait changer si d'autres recherches étaient menées. Comme peu d'études ont comparé les trois médicaments les uns aux autres, le fait qu'il n'y ait pas de différence entre eux en termes d'efficacité nous semble moyennement certain.

Notes de traduction

Traduction réalisée par le Centre Cochrane Français

Background

Description of the condition

Genital herpes is the most frequent cause of genital ulcer disease (Celum 2004). Genital herpes is a sexually transmitted disease caused by herpes simplex virus type 2 (HSV-2) in 60% to 80% of cases, but HSV-1-related genital herpes is increasing and accounts for about half of new cases in developed countries (Gupta 2007; Kimberlin 2004). Serological studies have found a seroprevalence of HSV-2 of 15% to 25% in the general population of developed countries (Looker 2008; Whitley 2001). The HSV-2 seroprevalence in people 14 to 49 years of age in the United States was 16.2% in 2005-2008, and did not significantly differ from that reported in 1999-2004 (MMWR 2010). Seroprevalence was highest among women (20.9%). Genital herpes favors transmission of human immunodeficiency virus (HIV) and other sexually transmitted diseases (Freeman 2006). Besides, transmission during childbirth can lead to neonatal herpes (Kimberlin 2013).

Infection occurs during close contact with the mucous membrane, abraded skin lesions or mucosal secretions of a person who has genital lesions or is shedding HSV. Viral invasion of epithelial cells occurs at the site of the exposed infection then HSV ascends to the sensory nerve root ganglia and enters a latent state. The primary HSV genital infection can be asymptomatic or characterized by the appearance four to seven days after sexual contact of painful mucocutaneous vesicles on the genital area, which evolve into erosions and crusts. Systemic symptoms such as fever may be associated (Fife 2006).

Recurrent episodes are caused by HSV reactivation. Viruses travel along sensory neurons to the corresponding mucocutaneous area (Gupta 2007). Reactivation of latent HSV leads to subclinical (asymptomatic shedding) or symptomatic genital mucocutaneous outbreaks. Recurrences are milder than primary infection, with more limited and less painful unilateral lesions and without systemic symptoms. Spontaneous healing occurs in five to 10 days (Sen 2007). Symptomatic recurrent flares occur in 20% to 50% of patients with anti-HSV antibodies. Following a symptomatic first episode of HSV-2 genital infection, the median recurrence rate is four recurrences during the first year (Sen 2007). The rate of recurrence usually decreases over time but increases in about one quarter of patients. Immunosuppressed patients have more severe and frequent recurrences (Benedetti 1994). Genital HSV-1 infection leads to less frequent outbreaks (mean recurrence rate 1.3/y) than are seen with HSV-2 infection (Engelberg 2003).

Shedding (viral detection at mucosal sites) occurs during symptomatic episodes of genital herpes but also in the absence of genital lesions, even in people without genital herpes recurrences. A prospective study observed 498 immunocompetent HSV-2-seropositive patients for a median of 57 days. HSV-2 viral shedding (measured by polymerase chain reaction (PCR)) was detected on 20% of days with swabs collected in people with symptomatic genital HSV-2 infection and on 10.2% of days in people with asymptomatic infection (Tronstein 2011). A meta-analysis found that people who reported always using condoms had a 30% decreased risk of HSV-2 infection as compared with people who reported no condom use (Martin 2009).

The clinical diagnosis of genital herpes and virus typing should be confirmed by laboratory testing. Viral isolation in cell culture has low sensitivity and allows virus typing and antiviral sensitivity testing. PCR, a rapid and sensitive method of viral detection and virus typing, is now frequently used (Scoular 2002). In a study of detection of HSV by PCR or by isolation in Vero cell culture on genital swabs taken from 194 participants who presented with genital ulceration or symptoms suggestive of genital herpes infection, comparison of cell culture versus PCR showed a sensitivity of 93/115 (80.9%) and a specificity of 79/79 (100%) (Slomka 1998).

HSV type-specific serology (enzyme-linked immunosorbent assay (ELISA) or Western-blot analysis) is indicated in the absence of or negative results of direct virus detection methods in patients with a history of recurrent or atypical genital lesions. HSV-2 antibodies suggest a diagnosis of genital herpes; the presence of HSV-1 antibodies does not allow differentiation between genital and oro-pharyngeal infection (European guidelines (IUSTI)).

Description of the intervention

Three oral antiviral drugs are available to treat genital herpes: acyclovir; valacyclovir, a prodrug of acyclovir with higher bioavailability after oral administration, and famciclovir (FCV). Famciclovir is the well-absorbed (77% bioavailable) oral form of penciclovir.

Randomized controlled trials demonstrated that these three drugs reduce the severity and duration of the recurrence by one to two days when started early after the beginning of the first symptoms (prodromes) or clinical signs of recurrence. This intermittent regimen does not reduce the number of recurrences by year (Abudalu 2008; Aoki 2006; Bodsworth 2009; Goldberg 1986;Leone 2002; Reichman 1984;Romanowski 2000; Sacks 2005; Saiag 1999 Tyring 1998; Wald 2002; Warren 2004).

Because none of these drugs can eradicate the latent virus, and because recurrent genital herpes is associated with medical and psychosocial morbidities, suppressive therapy is recommended in patients with frequent and/or severe episodes (Gupta 2007; Sen 2007). One trial demonstrated that valacyclovir as suppressive antiviral treatment decreased the risk of HSV-2 in serodiscordant couples (Corey 2004). Two randomised trials found that acyclovir suppressive therapy used to suppress HSV-2 failed to reduce the risk of HIV acquisition in HSV-2-infected but HIV-negative participants (Celum 2004; Watson-Jones 2008).

How the intervention might work

The action mechanism of these drugs consist of the following: viral thymidine kinase converts penciclovir or acyclovir to penciclovir or acyclovir monophosphate, which is converted by cellular enzymes to penciclovir or acyclovir triphosphate. Penciclovir or acyclovir triphosphate inhibits viral DNA polymerase, thereby halting DNA synthesis and viral replication (De Clercq 2006). The minimum inhibitory concentration of acyclovir is slightly higher for HSV-2 than for HSV-1 (Rosenberry 1982).

The suppressive antiviral therapy consists of daily oral administration of acyclovir, famciclovir, or valacyclovir for six to 12 months. After completion of the treatment regimen, the number of recurrences without suppressive therapy is assessed. The aim of the suppressive antiviral therapy is to reduce the number of recurrences, viral shedding, and the risk of transmission during treatment. Acyclovir was the first molecule to have a meaningful therapeutic effect in this setting, with significant efficacy (reduction of recurrence varying in clinical trials from 41% to 86% vs placebo). More recently, valacyclovir and famciclovir have been evaluated for this indication.

Why it is important to do this review

Most trials assessing antiviral drugs for HSV (at different doses and regimens) used a placebo or no treatment arm. A previous meta-analysis showed that acyclovir, famciclovir, and valacyclovir were superior to placebo or no treatment for prophylaxis of recurrent genital herpes (Lebrun-Vignes 2007). However, direct comparisons between the different regimens are lacking, and no conclusive evidence on the optimal antiviral therapy has been provided. It is important to synthesize up-to-date evidence from randomized controlled trials (RCTs), by using a network meta-analysis, and to compare the different drugs against one another, thus allowing for comparisons not addressed in the individual primary trials.

Objectives

To compare the effectiveness and safety of three oral antiviral drugs (acyclovir, famciclovir, and valacyclovir) prescribed to suppress genital herpes outbreaks versus placebo or versus one another in nonpregnant patients.

Methods

Criteria for considering studies for this review

Types of studies

Eligible studies were parallel-group or cross-over RCTs. No restriction was placed on publication status or language.

Types of participants

We reviewed studies of  participants with recurrent genital herpes (defined as four or more outbreaks per year) due to HSV, whatever the type (HSV-1, HSV-2, or undetermined). We identified a threshold that corresponded to an indication for suppressive therapy. See "Overall completeness and applicability of evidence" for discussion of the choice of four recurrences per year. Studies of HIV-positive participants or pregnant women were not eligible (Hollier 2008).

Types of interventions

Suppressive oral antiviral treatment with oral acyclovir, famciclovir and valacyclovir for genital herpes compared against each one other or compared with placebo. A suppressive treatment is defined a continuous treatment for several months that aims to prevent herpes recurrence. In multi-arm trials, study groups assessing drugs other than acyclovir, famciclovir or valacyclovir and study groups assessing intermittent treatment only were not eligible.

Types of outcome measures

Primary outcomes
  • Proportion of participants with at least one recurrent episode of genital herpes during the treatment period.

  • Recurrence-free survival.

Secondary outcomes
  • Proportion of participants with at least one virologically confirmed recurrent episode of genital herpes during the treatment period.

  • Rate of days with symptomatic or asymptomatic viral excretion (shedding) during the treatment period.

  • Proportion of participants with adverse events.

Search methods for identification of studies

Electronic searches

We searched a range of bibliographical databases, including the Cochrane Central Register of Controlled Trials (CENTRAL); the specialized registers of the Cochrane Infectious Diseases Group, the Cochrane Skin Group, and the Cochrane Sexually Transmitted Infections Group; MEDLINE, EMBASE and LILACS, with no restriction on language or date. The search was performed in collaboration with the Trials Search Co-ordinator of the Cochrane Sexually Transmitted Diseases Group and a healthcare librarian. MEDLINE and EMBASE were searched with both controlled vocabulary (namely, Medical Subject Headings (MeSH) in MEDLINE and EMTREE in EMBASE) and a wide range of free-text terms (Appendix 1, Appendix 2, Appendix 3). The search of MEDLINE involved use of the Cochrane Highly Sensitive Search Strategy, sensitivity-maximizing version. Another search equation included the filter used by the Cochrane HIV/acquired immunodeficiency syndrome (AIDS) group. The search of EMBASE involved use of the UK Cochrane Centre search filter to identify reports of randomized trials. The search last ran until February 2014.

Searching other resources

We screened the reference lists of all selected trials. We contacted the main authors in the field to identify any additional published or unpublished data. We searched the proceedings of the following conferences.

  • European Congress International Union Against Sexually Transmitted Infections (IUSTI), 2004, 2006, 2007, 2008, 2010, 2012, 2013

  • European Congress of Clinical Microbiology and Infectious Diseases, 2000-2013.

  • Interscience Conference on Antimicrobial Agents and Chemotherapy, 2008-2010.

  • American Academy of Dermatology, 2004-2013.

  • Annual Society for Dermatological Research Meeting, 2000-2013.

  • World Congress of IUSTI, 2007.

We contacted the pharmaceutical companies producing the antiviral agents of interest and searched the clinical trial results database of each company (www.novctrd.com and www.gsk-clinicaltrialregister.com) to identify ongoing and unpublished trials. We searched reviews submitted to the Food Drug Administration (FDA) for drug registration (http://www.accessdata.fda.gov/scripts/cder/drugsatfda/). We also searched the search portal of the World Health Organization (WHO) International Clinical Trials Registry Platform to identify ongoing trials. Finally, we sent the listing of included trials to two main authors in the field (Prof. Corey and Prof. Wald), and we asked whether they were aware of any additional potentially relevant trials.

The search of these sources was run until February 26, 2014.

Data collection and analysis

Selection of studies

Two review authors (LLC and GD) independently examined each title and abstract to exclude irrelevant reports. The two review authors independently examined full-text articles to determine eligibility. Trial authors were contacted for clarification when required to assess eligibility. Disagreements were discussed to reach consensus. We linked together multiple reports of the same study and we specified which report was used as the primary data source. We listed excluded trials and documented the primary reason for exclusion.

The trial selection process involved the use of Cochrane-resyweb.net, which handles the importation of references from multiple databases, as well as the semi-automatic deletion of duplicate records of the same report, selection of trials by independent review authors, the use of consensus procedures in cases of disagreement, and the linkage of multiple reports of the same trial. Finally, the web service automatically produced a flow chart for reporting the search process.

Data extraction and management

The data from published and unpublished reports were extracted independently by two review authors (LLC, and AM). They use a standardized form which was piloted. Consensus was achieved by discussion with a statistician (LT). We abstracted information regarding trial design and setting (parallel-arm or cross-over design, single center or multicenter, country, period of inclusion), population (age, sex, number of recurrences the previous year), interventions (dosage, frequency, duration, co-intervention), and outcome data. These data were checked and entered into the Cochrane Review Manager (RevMan) computer software by one review author (AM).

Assessment of risk of bias in included studies

The Cochrane Collaboration Risk of bias tool was used to assess the risk of bias. Randomization was considered adequate if the allocation sequence was generated from a table of random numbers or by computer. It was considered inadequate if sequences could be related to prognosis. It was considered unclear if it was stated that the report noted that the trial was randomized, but the method was not described. Allocation concealment was deemed adequate if the report stated that it was undertaken by means of sequentially pre-numbered, sealed opaque envelopes or by a centralized system. The risk of bias associated with inadequate blinding was evaluated separately for personnel and participants on one hand, and for outcomes assessors on the other hand. A double-blind double-dummy process was considered at low risk of bias. Concerning the risk of bias associated with incomplete outcome data, we examined the presence of imbalance in numbers or reasons for missing data across intervention groups, and determined wether the analysis was performed according to the intention-to-treat principle. The use of strategies to handle missing data (last observation carried forward, multiple imputation, etc.) was assessed. For selective reporting of outcomes, we considered the risk of bias high when results for outcomes detailed in the methods section were not reported in the results section; otherwise, the risk of bias was considered as unclear in the absence of an available protocol. A risk of bias table was completed for each included trial. Moreover, for cross-over trials, we took into account the following items according to Section 16.4.3 of the Cochrane Handbook for Systemic Reviews of Interventions (Cochrane Handbook) : whether a carry-over effect was evident, whether only first period data were available, and whether the analysis performed was appropriate.

Measures of treatment effect

For at least one recurrence of genital herpes, we measured treatment effects using risk ratios (RRs) and associated 95% confidence intervals (95% CIs).

Unit of analysis issues

The primary unit of analysis was the participant. In parallel-group trials, participants are randomly assigned to one intervention. In cross-over trials, participants are randomly assigned to a sequence of interventions, and participants act as their own control. Cross-over studies that ignore the pairing commit a unit-of-analysis error. We have reported such errors, but we could not reanalyze data appropriately (Curtin 2002).

Dealing with missing data

We checked whether each trial used the intention-to-treat principle for analysis (i.e., the number of analyzed participants equals the number of randomly assigned participants). For the main analysis, we assumed that any participants with missing outcome data experienced at least one recurrent episode, whatever the group. We also synthesised data as analyzed in each trial (complete cases).

Assessment of heterogeneity

To assess to what extent the results of trials were consistent, we examined variability in point estimates and overlap of confidence intervals across studies; we used Higgins and Thompson's I2 statistic. We interpreted the I2 value according to the following thresholds (Section 9.5.2 of the Cochrane Handbook): 0% to 40% might not be important; 30% to 60% may represent moderate heterogeneity; 50% to 90% may represent substantial heterogeneity; and 75% to 100% represents considerable heterogeneity. We also computed the 95% CI for the I2 (Ioannidis 2007). In all cases, we considered the results from both the fixed-effect model and random-effects models, and the choice between the two models was based on the number of studies, the distribution of effect sizes, measures of heterogeneity, and clinical and methodological diversity.

Assessment of reporting biases

To address reporting biases and related small-study effects, we drew funnel plots for each pairwise meta-analysis. Funnel-plot asymmetry tests were used when the validity conditions were met: no significant heterogeneity (I2 statistic < 50%), 10 or more studies and at least one with significant results, and a ratio of maximal to minimal variance across studies > 4 (Ioannidis 2007a). In cases of evidence of small-trial effects, we performed sensitivity analyses according to a regression-based adjustment model (Moreno 2009;Rucker 2011).

Data synthesis

First, we derived pooled RR estimates from pairwise meta-analyses for available direct comparisons. The Mantel-Haenszel method was used. We analyzed cross-over trials by a simple approach using all outcome data from the experimental intervention periods and all outcome data from the control intervention periods and analyzing these as if the trial were a parallel-group trial. This incorrect analysis is conservative. One trial reported data from the first period only, so we analyzed only these data. Cross-over trials were synthesized separately from parallel-group trials and in cases of differences between the two strata, we did not combine all trials together. In cases of multi-arm trials, we compared arms two at a time in separate comparisons. In cases of multi-dose trials, we grouped together all the different dose groups and compared them collectively with the control group.

Second, we derived pooled estimates for all possible two-by-two comparisons between interventions from a network meta-analysis; a hierarchical model with a Bayesian approach was used (Lu 2004). This model handles multi-arm trials appropriately. We assessed the consistency of the network by using inconsistency factors on closed loops; we estimated the difference between direct and adjusted indirect estimates and associated 95% CIs and tested whether it was statistically significant. Estimates of three two-by-two comparisons between each antiviral drug and the others were reported, along with 95% credibility intervals (95%CrIs). The rank-order of efficacy of the different drugs and the probability that each drug is the best among the three antiviral drugs were estimated (Salanti 2010).  We also performed a sensitivity analysis to adjust for small-study effects based on a network meta-regression model, which assumed that biases were exchangeable across comparisons versus placebo across the network (Trinquart 2012).

Pairwise meta-analyses involved use of RevMan software, adjustment models for small-study effects were fitted by the use of R software, and network meta-analyses involved the use of WinBUGS software. Finally, we presented the findings into a "Summary of findings" table including a summary of the amount of evidence, estimates of relative effects of treatments and estimates of typical absolute risks for people receiving drugs or placebo. We also assessed the quality of the body of evidence according to the GRADE framework .

Subgroup analysis and investigation of heterogeneity

in order to assess the influence of doses, we performed subgroup analyses according to the total daily dose. Subgroup analyses initially planned to determine efficacy in men versus women, with HSV-1 versus HSV-2 genital herpes, were not possible because these data were not reported in the included trials.

Results

Description of studies

Results of the search

Overall, we included 26 trials corresponding to 42 reports (Figure 1). The search of electronic databases (MEDLINE, EMBASE, LILACS, and CENTRAL) yielded 5563 citations. The search of other sources identified nine reports: we found eight reports through the clinical trial results databases of pharmaceutical companies (four were also published (Bartlett 2008; Fife 2006; Reitano 1998; Patel 1997), and one report was identified by screening conference proceedings (Sekhin 2004)). No report was identified by searching FDA reviews and the WHO International Clinical Trials Registry Platform nor by contacting the main authors in the field. The two main authors in the field whom we contacted were not aware of any other relevant trial.

Figure 1.

Flow of information through the different phases of t he review

After duplicate records were removed, we had 3785 citations. After reviewing the titles and abstracts, we discarded 3676 citations. We examined the full text of the remaining 109 citations: 61 did not meet the inclusion criteria (see Excluded studies), and six trials were classified as studies awaiting classification. Among these, information regarding risk of bias assessment and data analyses were lacking in four trials ( Douglas 1988; Frenkel 1989; Mastrolorenzo 1991; Tyring 2003). We contacted the study authors, but they did not provide information (Table 1). For two trials potentially eligible according to the abstract, we could not retrieve the full-text article (Wu 2000, Zhang 2011).

Table 1. Details of contacting authors
StudyContactRequested informationContactedReply (last check 1/21/2013)  
Mastrolorenzo 1991Pr Mastrolorenzo

Outcome : participants with at least one recurrence

Outcome: time to first recurrence

 

October 6, 2012

October 17, 2012

Additional data to the publication not provided

PREV123 2006

HS240017 2005

GSK pharmaceutical company

PREV123 outcome : participants with at least 1 recurrence

HS240017 outcome :participants with at least 1 recurrence

HS240017 outcome: time to first recurrence

 

October 6, 2012

 

PREV 123 detailled report sent

 

HS240017 outcome data not collected during this study

Sekhin 2004Pr Stratchounski

Outcome: time to first recurrence

 

October 6, 2012

October 17, 2012

Additional data to the publication not provided
Douglas 1988CDC (Pr Douglas)

Outcome : participants with at least 1 recurrence

Outcome: time to first recurrence

 

October 6, 2012

October 17, 2012

No response
Mattison 1988Pr Corey

Outcome: participants with at least 1 recurrence

Outcome: time to first recurrence

 

October 6, 2012

October 17, 2012

No response
Frenkel 1989Pr Bryson

Outcome: participants with at least 1 recurrence

Outcome: time to first recurrence

 

October 6, 2012

October 17, 2012

No response
Wald 2006 study 2Pr Wald

Outcome: participants with at least 1 recurrence

Outcome: time to first recurrence

 

October 6, 2012

 

Data no longer available
Tyring 2003Pr TyringArticle reported the pooled results of 2 trials, among which one was already published (Diaz-Mitoma 1998). We asked for the data from the other trialJanuary 12, 2012No response
Corey 2004bPr CoreyOutcome data for the subgroup of participants with more than 4 recurrences per yearJanuary 16, 2012Data no longer available

Among the 26 included trials, three trials (Fife 2006; Patel 1997; Reitano 1998) provided both published and unpublished results; when this occurred, we used the unpublished results because numerical data were available (for two trials (Fife 2006;Patel 1997) discrepancies were observed). Results of six trials were published in more than one journal article (Halsos 1985; Mattison 1988; Mertz 1988; Reitano 1998; Romanowski 2003; Straus 1984). For two trials, only unpublished results were available (HS240017 2005; PREV123 2006).

See Characteristics of included studies

In all, 22 studies declared pharmaceutical company funding, and four studies did not report the source of funding (Blom 1986; Kroon 1989; Mindel 1984; Sekhin 2004). In 13 of the 22 studies, at least one of the trial authors was employed by the funding pharmaceutical company.

Trial design

A total of 20 trials used a parallel-group design, and six a cross-over design (Blom 1986; Halsos 1985; Kinghorn 1985; Kroon 1989; Romanowski 2003; Thin 1985). Across the 26 trials, the median sample size was 95 (Table 2). For the parallel-group trials, the median sample size was 128, and for cross-over trials, 40. In all, 16 trials were multicenter trials (two to 76 centers) and eight were single-center trials (Douglas 1984; Kinghorn 1985; Kroon 1989; Mindel 1984; Mostow 1988; Sacks 1988; Sacks 1988; Velasco 1991); for two trials, single-center or multicenter status was not clear (Sekhin 2004; Straus 1984). Four multiarm trials assessed the same experimental drug at multiple dose levels (acyclovir in Douglas 1984; valacyclovir in Reitano 1998; famciclovir in Diaz-Mitoma 1998 and Mertz 1997). Moreover, Reitano 1998 assessed two experimental drugs (acyclovir and valacyclovir) versus placebo.

Table 2. Characteristics of participants summarized across RCTs
  1. ACV: acyclovir, FCV: famciclovir; VCV: valacyclovir; PBO: placebo; No trt: no treatment

    *Data are reported as median (range) for the number of randomly assigned participants

 ALL COMPARISONSACV vs PBOFCV vs PBOVCV vs PBOVCV vs No trtACV vs VCVFCV vs VCV
 # RCTs (n=26)Mean (range)# RCTs (n = 15)Mean (range)# RCTs (n = 3)Mean (range)# RCTs (n = 5)Mean (range)# RCTs (n = 2)Mean (range)# RCTs (n = 1)Mean (range)# RCTs (n = 2)Mean (range)
Publication year261992 (1984-2006)151990 (1984-1998)31998 (1997-2004)52004 (1997-2006)22005 (2003-2006)1199822006 (2006-2006)
No of participants*2695 (29-1479)1555 (29-1175)3

375

(180-457)

5382 (42-1479)2

368

(225-511)

1

1345

(NA)

2195 (70-320)
Age, years24

35.2

(26.0-45.1)

14

32.9

(26.0-38.0)

3

35.0

(32.5-36.8)

4

34.3

(33.2-40.1)

2

42.9

(38.0-45.1)

1

33.8

(NA)

2

36.7

(35.4-37.0)

Female25

54.0%

(0%-100%)

15

42.4%

(0-66.7%)

3

77.9%

(50.1%-100%)

4

57.1%

(41.0%-75%)

2

61.5%

(61.4%-61.6%)

1

48.6%

(NA)

2

60.3%

(52.9%-60.6%)

No of recurrences the previous year13

11.6

(5.9-17.8)

11

12.9

(11-17.8)

0NA0NA1

5.9

(NA)

0NA2

9.2

(9.1-9.7)

Characteristics of participants

The characteristics of participants are summarized in Table 2. Across the 26 included trials, 1437 participants were randomly assigned to receive acyclovir, 2123 valacyclovir, and 1162 famciclovir (Table 2). Moreover, 1659 participantss were randomly assigned to receive placebo and 569 no treatment. Females represented 54% (min to max 0% to100%) of included participants, Halsos 1985 included only men, and Mertz 1997 and Sacks 2004 included only women. The overall mean age of participants was 35.0 (min to max 26.0 to 45.1) years. The overall mean number of recurrences was 11.0 (range 6.3 to 17.8) (information available for 15 trials). For 15 trials, confirmation of genital herpes by viral culture or PCR was required for inclusion, and in the other trials, a clinical diagnosis of genital herpes without laboratory confirmation was sufficient. In four trials, all participants had recurrent genital herpes due to HSV2 (Blom 1986; Douglas 1984; Fife 2006; Sacks 2004). The number of participants with HSV-1 or HSV-2 infection was reported in one study (Kinghorn 1985). For the others, this information was not provided.

Comparisons

A total of 14 trials compared acyclovir with placebo (Blom 1986; Douglas 1984; Halsos 1985; Kinghorn 1985; Kinghorn 1992; Kroon 1989; Mattison 1988; Mertz 1988; Mindel 1984; Mostow 1988; Sacks 1988; Straus 1984; Thin 1985; Velasco 1991). Four trials compared valacyclovir with placebo (Fife 2006; HS240017 2005; Patel 1997; Sekhin 2004) and two trials compared valacyclovir with no treatment (PREV123 2006; Romanowski 2003). Three trials compared famciclovir with placebo (Diaz-Mitoma 1998; Mertz 1997; Sacks 2004). Two trials compared valacyclovir with famciclovir (Wald 2006 study 1; Wald 2006 study 2) and one trial compared acyclovir with valacyclovir with placebo (Reitano 1998).

Characteristics of interventions

The duration of suppressive treatment was two months in two trials (Fife 2006, HS240017 2005), 2.5 months in one (Wald 2006 study 2), three months in six (Blom 1986; Halsos 1985; Kinghorn 1985; Kroon 1989; Mindel 1984; Thin 1985), four months in seven (Douglas 1984; Patel 1997; Reitano 1998; Sacks 2004; Sekhin 2004; Straus 1984; Wald 2006 study 1), six months in five (Kinghorn 1992; PREV123 2006; Romanowski 2003; Sacks 1988; Velasco 1991), and 12 months in five (Diaz-Mitoma 1998; Mattison 1988; Mertz 1988; Mertz 1997; Mostow 1988). For five studies, treatment was stopped after the first recurrence (Blom 1986; Kinghorn 1985; Patel 1997; Straus 1984; Thin 1985). The suppressive treatment duration for cross-over trials was three months, except for one of six months' duration (Romanowski 2003). For cross-over trials, the wash-out period was never indicated; however, it was specified in three studies that second period treatment was started after a new recurrence occurred ( Blom 1986; Kinghorn 1985; Thin 1985).

In 12 trials, the daily dose received by participants randomised to receive acyclovir was 800 mg per day distributed as follows: 200mg 4 times a day in 5 trials ( Mindel 1984; Blom 1986; Halsos 1985; Kinghorn 1985; Thin 1985), 400 mg twice a day in 6 trials (Kroon 1989; Reitano 1998; Mertz 1988; Mattison 1988; Kinghorn 1992; Velasco 1991) and 800 mg once a day in 1 trial (Mostow 1988); it was 600 mg per day (200 mg 3 times a day) in 2 trials (Sacks 1988; Straus 1984). There was one multi-dose trial: 400 mg per day (200mg twice a day) versus 1000 mg per day (200mg five times a day) versus placebo (Douglas 1984).

The daily dose received by participants randomly assigned to receive valacyclovir was 1000 mg per day (500 mg twice daily) in one trial (Fife 2006). In seven trials, it was 500 mg per day, distributed as follows: 500 mg once a day in six trials (HS240017 2005; Patel 1997; PREV123 2006 for participants with < 10 recurrences per year; Romanowski 2003; Wald 2006 study 1; Wald 2006 study 2) and 250 mg twice a day in one trial (PREV123 2006 for patients with > 10 recurrences per year). One multidose trial provided 1000 mg once a day versus 250 mg twice a day versus 500 mg once a day versus 250 mg once a day versus acyclovir versus placebo (Reitano 1998).

The daily dose received by participants randomly assigned to receive famciclovir was 250 mg twice daily in two trials (Wald 2006 study 1; Wald 2006 study 2). Three mutidose trials provided 250 mg three times daily versus 250 twice a day versus 125 mg three times daily versus placebo (Diaz-Mitoma 1998); 500 mg once daily versus 250 mg twice daily versus 250 mg once daily versus 125 mg twice daily versus 125 mg once daily versus placebo (Mertz 1997); and 250 mg three times daily versus 125 mg twice daily versus placebo (Sacks 2004).

In nine trials, participants received intermittent antiviral treatment for recurrence (Fife 2006; Halsos 1985; HS240017 2005; Mertz 1988; Mostow 1988; PREV123 2006; Reitano 1998; Romanowski 2003; Sacks 1988). In two trials, participants received the intermittent treatment in one arm and suppressive treatment in the other arm (Bartlett 2008; Fife 2007).

Excluded studies

We classified six trials as studies awaiting classification, and we excluded 61 full-text reports. A total of 28 reports were not RCT reports, and in 11 trials, the experimental treatment consisted of intermittent treatment only. The reason for exclusion of the 22 other reports (corresponding to 19 trials) are detailed in Characteristics of excluded studies. The main reason for exclusion (10 reports corresponding to eight studies) was the absence of a minimum number of recurrences required for inclusion. For one trial (Corey 2004), we asked the study authors to provide the results for the subgroup of participants with at least four recurrences, but these data were not available and the study was excluded (Table 1). For two studies with three arms, one arm was not included: Kinghorn 1992 compared acyclovir versus placebo versus isoprinosine, and the isoprinosine arm was not included; Mattison 1988 compared acyclovir plus placebo intermittent treatment versus placebo acyclovir plus placebo intermittent treatment versus placebo acyclovir plus intermittent treatment, and the last arm was not included. Finally, two trials identified from searches of clinical trial results databases of pharmaceutical companies were excluded because they were stopped after two and six inclusions, respectively, following a decision by the FDA (HS240018; HS240021).

Risk of bias in included studies

Risk of bias assessments are summarized in Figure 2 and Figure 3. Regarding the overall risk of bias, no trial had low risk of bias, 13 were at high risk of bias, and 13 were at unclear risk of bias.

Figure 2.

Risk of bias summary: review authors' judgements about each risk of bias item for each included trial.

Figure 3.

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Allocation

In all, 22 trials were reported as randomized, but as reports did not contain a description of the process of sequence generation, they were considered at unclear risk of bias for generation of sequence generation. For the four other trials, the process of sequence generation was reported and adequate, and these studies were considered at low risk of bias for sequence generation (Diaz-Mitoma 1998; Kinghorn 1992; Mostow 1988; Romanowski 2003). For allocation concealment, the risk of bias was considered unclear in 24 trials because of absence of reporting of the method used to guarantee concealment. The risk was considered low for two trials (Diaz-Mitoma 1998; Fife 2006).

Blinding

Two trials were open-label (PREV123 2006; Romanowski 2003). Given that the outcome was subjective, we considered the risk of bias as high for these studies. A total of 24 trials were reported as double-blind. These studies were placebo-controlled or double-dummy. We thus considered these trials at low risk of bias, even if no additional information was available.

Incomplete outcome data

Seven trials were considered at high risk of bias because of the high numbers of withdrawn participants and/or because of an imbalance between groups in the numbers of withdrawn participants and/or an imbalance in reasons for missing outcomes (Diaz-Mitoma 1998; Mattison 1988; Mertz 1988; Mostow 1988;Patel 1997; PREV123 2006; Romanowski 2003). Six trials were considered at unclear risk of bias because the numbers of participants and/or reasons and/or missing data methods were not reported.

Selective reporting

Five trials were considered at high risk of selective outcome reporting because results for outcomes detailed in the methods section were not reported in the results section. For HS240017 2005 duration of recurrences and duration of clinical shedding, and for Kinghorn 1985; Mattison 1988; Mostow 1988; and Thin 1985, number of recurrences, were confirmed by positive viral culture. For other studies, the risk of bias was considered unclear because we did not have access to any protocol.

Other potential sources of bias

In the cross-over design, no mention was made in any trial of the duration of the washout period and of the possibility of a carry-over of treatment effect across periods; we considered that the carry-over effect was unlikely because of the natural course of the disease, and because the half-life of the drugs is short. However, all cross-over trials had unit of analysis errors. In one trial, results were reported for the first period only (Romanowski 2003), and the trial was considered at high risk of bias. In the other cross-over trials, the analysis was incorrect: in two trials, results were reported only from the experimental and control treatment periods combined together (Blom 1986; Halsos 1985). In other cross-over trials, results were reported for the two treatment periods, but paired data were not reported, so adequate analysis was not possible (Curtin 2002).

Effects of interventions

At least one clinical recurrence (pairwise meta-analyses)

We were not able to analyze this outcome for four included trials (HS240017 2005; Mattison 1988; Sacks 2004; Wald 2006 study 2). The outcome was not reported, and we were not able to obtain missing information from the trial authors (Table 1). For one trial (PREV123 2006), information was provided after contact was made with the pharmaceutical company. For another trial (Sacks 2004), only virologically confirmed recurrences were reported. The findings are presented in the Summary of findings table (Table 3).

Table 3. Summary of findings

participantswith recurrent genital herpes and at least 4 recurrences per year

Intervention: suppressive antiviral therapy

Comparison: placebo

Outcome: at least 1 clinical recurrence

 AcyclovirValacyclovirFamciclovirComments
As compared with placebo

RR 0.48

(0.39-0.58)

RR 0.41

(0.24-0.69)

RR 0.57

(0.50-0.64)

The network meta-analysis of the few direct comparisons and the indirect comparisons did not show superiority of one drug over another. We did not include the results of the network meta-analysis in this table because the methods for rating the quality of evidence for network meta-analyses are not well-established.
90 per 100 participants (90%)

47 fewer per 100

(38-55)

53 fewer per 100

(28-68)

39 fewer per 100

(32-45)

 

Low

confidence in estimate due to heterogeneity, risk of bias and small-study effect

based on 2049 participants (9 trials)

Low

confidence in estimate due to heterogeneity and risk of bias

based on 1788 participants (4 trials)

Low

confidence in estimate due to heterogeneity and risk of bias

based on 732 participants (2 trials)

Acyclovir versus placebo

Fourteen trials (2549 randomly assigned participants) were analyzed (Figure 4). In the nine parallel-group trials (2049 randomly assigned participants), 1462 participants had at least one recurrence: 583 in the acyclovir arm and 879 in the placebo arm. The risk of at least one recurrence was reduced with acyclovir rather than placebo (pooled RR 0.48, 95% CI 0.39 to 0.58). Substantial heterogeneity was noted across trials (I² = 81%, 95% CI 64% to 90%). In the five cross-over trials (500 participants), reported analyses did not take into account the pairing. We analyzed them through a simple approach and found no evidence of statistical heterogeneity (I² = 0, 95% CI 0% to 79%); however, the pooled treatment effect estimate was significantly larger than in parallel-group trials (pooled RR 0.35, 95% CI 0.29 to 0.42), so we did not combine cross-over trials with parallel-group trials.

Figure 4.

Forest plot of comparison: 1 Acyclovir vs placebo, outcome: 1.1 Participants with at least 1 clinical recurrence.

The funnel plot showed strong asymmetry, with smaller trials showing larger treatment effect estimates than larger trials (Figure 5).

Figure 5.

Funnel plot of comparison: 1 Acyclovir vs placebo, outcome: 1.1 Participants with at least 1 clinical recurrence.

For the nine parallel-group trials, we performed a sensitivity analysis for small-study effects: when we used the regression-based approach of Moreno et al, the pooled treatment effect adjusted for small-study effects still showed a significant benefit with acyclovir but was smaller than with the unadjusted analysis (pooled RR 0.61, 95% CI 0.49 to 0.75). When the limited meta-analysis of Rucker et al was used, findings were similar (pooled RR , 0.63, 95% CI 0.54 to 0.73).

When we analyzed the nine parallel-group trials by daily dose, we found no evidence of a clear dose-response relationship (test for subgroup differences P value 0.38); (Figure 6). The comparison between acyclovir 800 mg/d and placebo involved the largest number of participants (pooled RR 0.52, 95% CI 0.43 to 0.64). The other daily doses involved one randomly assigned comparison only (400 and 1000 mg/d) and three randomly assigned comparisons (totaling 140 participants) regarding acyclovir 600 mg/d; in the latter case, heterogeneity was substantial and the pooled estimate may not be relevant.

Figure 6.

Forest plot of comparison: 1 Acyclovir vs placebo, outcome: 1.2 Participants with at least one clinical recurrence (according to daily dose).

Results from the analysis of data without imputation of missing outcomes (complete cases as analyzed in trials) were similar to results from our main analysis, although treatment effect estimates were slightly larger (Figure 7).

Figure 7.

Forest plot of comparison: 1 Acyclovir vs Placebo, outcome: 1.3 Participants with at least 1 clinical recurrence (complete cases).

Valacyclovir versus placebo or no treatment

Four parallel-group trials (1788 randomly assigned participants) compared valacyclovir versus placebo (Figure 8). The risk of at least one recurrence was reduced with valacyclovir rather than placebo (pooled RR 0.41, 95% CI 0.24 to 0.69). Substantial heterogeneity was noted across trials (I² = 94%, 95% CI 87% to 97%). Because of the small number of trials, the funnel plot did not allow for assessment of small-study effects. When we analyzed the four parallel-group trials by daily dose, we found moderate evidence of a dose-response relationship (Figure 9). However, heterogeneity across trials was noted. The comparison between valacyclovir 500 mg/d and placebo involved the largest numbers of participants. Note that for valacyclovir 1000 mg/d, heterogeneity was particularly substantial, and the pooled estimate may not be relevant.

Figure 8.

Forest plot of comparison: 3 Valacyclovir vs placebo, outcome: 3.1 Participants with at least 1 clinical recurrence.

Figure 9.

Forest plot of comparison: 2 Valacyclovir vs placebo, outcome: 2.2 Participants with at least 1 clinical recurrence (according to daily dose).

Two trials compared valacyclovir versus no treatment (Figure 10). One parallel-group trial (237 participants) showed the superiority of valacyclovir (RR 0.46, 95% CI 0.37 to 0.56) with consistent results as compared with trials versus placebo. One cross-over trial (450 participants) reported outcome data from the first period only (RR 0.57, 95% CI 0.50 to 0.66). We did not combine the findings of the cross-over trial with those of the parallel-group trial.

Figure 10.

Forest plot of comparison: 3 Valacyclovir vs No treatment, outcome: 3.1 Participants with at least 1 clinical recurrence.

Finally, results from the analysis of data without imputation of missing outcomes (complete cases as analyzed in trials) were similar to results from our main analysis, although treatment effect estimates were slightly larger (Figure 11; Figure 12).

Figure 11.

Forest plot of comparison: 2 Valacyclovir vs placebo, outcome: 2.3 Participants with at least 1 clinical recurrence (complete cases).

Figure 12.

Forest plot of comparison: 3 Valacyclovir vs No treatment, outcome: 3.2 Participants with at least 1 clinical recurrence (complete cases).

Famciclovir versus placebo

Two parallel-group trials (732 randomly assigned participants) compared famciclovir versus placebo (Figure 13). The risk of at least one recurrence was reduced with famciclovir rather than placebo (pooled RR 0.57, 95% CI 0.50 to 0.64). We found no evidence of statistical heterogeneity (I²= 0%). Because of the small number of trials, the funnel plot did not allow for assessment of small-study effects.

Figure 13.

Forest plot of comparison: 2 Famciclovir vs control, outcome: 2.1 Participants with at least 1 clinical recurrence.

When we analyzed the two parallel-group trials by daily dose, we found a tendency toward a dose-response relationship, but the test for subgroup differences revealed no significance (P value = 0.21 (Figure 14). Finally, results from the analysis of data without imputing missing outcomes (complete cases as analyzed in trials) were similar to results from our main analysis although treatment effect estimates were slightly larger (Figure 15).

Figure 14.

Forest plot of comparison: 4 Famciclovir vs placebo, outcome: 4.2 Participants with at least 1 clinical recurrence (according to daily dose).

Figure 15.

Forest plot of comparison: 4 Famciclovir vs placebo, outcome: 4.3 Participants with at least 1 clinical recurrence (complete cases).

Valacyclovir versus acyclovir

One parallel-group trial (1345 randomly assigned participants) compared valacyclovir versus acyclovir (Figure 16). The risk of at least one recurrence was increased with valacyclovir rather than acyclovir (RR 1.16, 95% CI 1.01 to 1.34). Because only one trial was found, the funnel plot did not allow for assessment of small-study effects.

Figure 16.

Forest plot of comparison: 4 Valacyclovir vs. Acyclovir, outcome: 4.1 participants with at least one clinical recurrence.

Finally, because the analysis in this trial was performed according to intention-to-treat principles, results from analysis of data without imputation of missing outcomes (complete cases as analyzed in trials) were similar to results from our main analysis (Figure 17).

Figure 17.

Forest plot of comparison: 5 Valacyclovir vs acyclovir, outcome: 5.2 Participants with at least 1 clinical recurrence (complete cases).

Famciclovir versus valacyclovir

One parallel-group trial (320 randomly assigned participants) compared famciclovir with valacyclovir (Figure 18). The risk of at least one recurrence was not significantly different with famciclovir rather than valacyclovir (RR 1.18, 95% CI 0.86 to 1.63). Because only one trial was identified, the funnel plot did not allow for assessment of small-study effects.

Figure 18.

Forest plot of comparison: 5 Famciclovir vs valacyclovir, outcome: 5.1 Participants with at least 1 clinical recurrence.

Finally, results from the analysis of data without imputation of missing outcomes (complete cases as analyzed in trials) were similar to results from our main analysis, although treatment effect estimates were slightly larger (Figure 19).

Figure 19.

Forest plot of comparison: 6 Famciclovir vs valacyclovir, outcome: 6.2 Participants with at least 1 clinical recurrence (complete cases).

At least one recurrence (network meta-analyses)

We included 16 parallel-arm trials in a network meta-analysis: eight trials compared acyclovir versus placebo ( Douglas 1984; Kinghorn 1992; Mertz 1988; Mindel 1984; Mostow 1988; Sacks 1988; Straus 1986; Velasco 1991), four trials compared valacyclovir versus placebo/no treatment (which were lumped together) (Fife 2006; Patel 1997; PREV123 2006; Sekhin 2004), two trials compared famciclovir versus placebo ( Diaz-Mitoma 1998 , Mertz 1997), one trial compared famciclovir versus valacyclovir (Wald 2006 study 1), and one trial compared acyclovir versus valacyclovir versus placebo (Reitano 1998) (Figure 20). Of the three possible pairwise comparisons between the three treatments, two were addressed directly, each by one trial.

Figure 20.

Network of randomly assigned comparisons from parallel-arm trials for the network meta-analysis of efficacy (at least 1 recurrence).

Convergence of the network meta-analysis model was reached. The three drugs were significantly superior to placebo in reducing the risk of at least one recurrence (pooled RR 0.46, 95% CI 0.34 to 0.57 for acyclovir, pooled RR 0.42, 95% CI 0.31 to 0.56 for valacyclovir; and pooled RR 0.57, 95% CI 0.37 to 0.84 for famciclovir). However, results were not statistically significantly different when drugs were compared against each other (pooled RR 0.94, 95% CI 0.65 to 1.38 for valacyclovir vs acyclovir; pooled RR 1.27, 95% CI 0.78 to 2.06 for famciclovir vs acyclovir; and pooled RR 1.37, 95% CI 0.86 to 2.11 for famciclovir vs valacyclovir). The probability of being the best antiviral drug was 29% for acyclovir, 65% for valacyclovir, and 6% for famciclovir (Figure 21), but we found substantial uncertainty in rank-order efficacy. In fact the ranks and associated 95% CrIs were second (first to third) for acyclovir, first (first to third) for valacyclovir, and third (first to third) for famciclovir. Of note, we found no evidence of loop inconsistency, although direct evidence suggested that valacyclovir was inferior to acyclovir, whereas indirect evidence indicated the contrary (inconsistency factors on the the log odds ratio scale: φ = 0.282, P value = 0.34 for valacyclovir-acyclovir-placebo; φ = -0.048, P value = 0.85 for famciclovir-acyclovir-placebo).

Figure 21.

Ranking for efficacy (at least 1 recurrence): probability to be the best treatment, the second, the third, or the fourth among acyclovir, valacyclovir, famciclovir, placebo/no treatment.

Estimates from the unadjusted analysis are in red; estimates from the adjusted analysis on small-study effects are in black.

In a sensitivity analysis of small-study bias, treatment effect estimates of the three drugs against placebo were reduced (pooled RR 0.59, 95% CI 0.47 to 0.73 for acyclovir, pooled RR 0.48, 95% CI 0.37 to 0.61 for valacyclovir; and pooled RR 0.60, 95% CI 0.44 to 0.82 for famciclovir). The adjustment in small-study effect had the largest impact on the comparison between acyclovir versus placebo, which was addressed by the largest number of trials. Treatment contrasts between drugs were slightly modified (pooled RR 0.83, 95% CI 0.61 to 1.11 for valacyclovir vs acyclovir; pooled RR 1.04, 95% CI 0.71 to 1.49 for famciclovir vs acyclovir; and pooled RR 1.26, 95% CI 0.89 to 1.75 for famciclovir vs valacyclovir), but, again, with no statistical difference noted. The probability of being the best antiviral drug was 7% for acyclovir, 85% for valacyclovir, and 8% for famciclovir; the ranks and associated 95% CrIs were second (first to third) for acyclovir, first (first to second) for valacyclovir, and third (first to third) for famciclovir, with overlap in 95% Crls.

At least one virologically confirmed recurrence (virological confirmation by viral culture)

Acyclovir versus placebo

Three parallel-group trials (138 randomly assigned participants) were analyzed (Figure 22). The risk of at least one virologically confirmed recurrence was reduced with acyclovir rather than placebo (pooled RR 0.08, 95% CI 0.03 to 0.22). No substantial heterogeneity was noted across trials (I² = 0%, 95% CI 0% to 90%).

Figure 22.

Forest plot of comparison: 1 Acyclovir vs placebo, outcome: 1.4 Particpants with at least 1 virologically confirmed recurrence.

Famciclovir versus placebo

Two parallel-group trials (555 randomly assigned participants) were analyzed (Figure 23). The risk of at least one virologically confirmed recurrence was reduced with famciclovir rather than placebo (pooled risk ratio 0.41, 95% CI 0.25–0.65). We found substantial heterogeneity across trials (I²= 76%).

Figure 23.

Forest plot of comparison: 4 Famciclovir vs placebo, outcome: 4.4 Participants with at least 1 virologically confirmed recurrence.

Famciclovir versus valacyclovir

One parallel-group trial (315 pooled RR ) was analyzed (Figure 24). The risk of at least one virologically confirmed recurrence was increased with famciclovir rather than valacyclovir (pooled RR 2.24, 95% CI 1.05 to 4.76).

Figure 24.

Forest plot of comparison: 6 Famciclovir vs valacyclovir, outcome: 6.3 Participants with at least one virologically confirmed recurrence.

Recurrence-free survival

Recurrence-free survival data were reported in 19 of 26 trials but varied substantially across trials: some trials reported hazard ratios derived from Cox models, others reported median time to the first recurrence in each treatment group, and some trials reported median time ratios. This variability did not allow for synthesis.

Viral shedding

Shedding was assessed in four trials. Daily genital swabs were analyzed in three trials by PCR (Fife 2006; HS240017 2005; Wald 2006 study 2) and in one by viral culture (Sacks 2004). Total number of swabs analyzed and/or number of missing swabs was not reported. These results did not allow for synthesis.

Adverse effects

The number of withdrawals due to harms in each arm was reported for only eight trials (Diaz-Mitoma 1998; Douglas 1984; Fife 2006; HS240017 2005; Mattison 1988; Mertz 1988; Patel 1997; Sacks 1988). Overall, 31 withdrawals due to harms occurred in the antiviral groups and 14 in the placebo or no treatment groups. Safety data were reported as total number of adverse events in only four trials (Fife 2006; HS240017 2005; Patel 1997; PREV123 2006). Overall, we found 331 adverse events, including three serious adverse events (one hypertension crisis, one intestinal obstruction, and one angor) across 561 participants in antiviral groups and 115 adverse events, including three serious adverse events (two renal signs and 1 fatal pneumonia) in 291 participants in the placebo or no treatment groups.

Discussion

Summary of main results

Overall, the results of our review show that acyclovir, famciclovir, and valacyclovir were all superior to placebo for the outcome of risk of at least one clinical recurrence of genital herpes while on treatment. On subgroup analysis of daily doses, we found no strong evidence of a dose-response relationship for any drug.

We found evidence of a small-study effect in the comparison of acyclovir versus placebo, with smaller trials showing larger treatment effect estimates than larger trials. We were not able to examine a potential small-study effect for valacyclovir versus placebo and famciclovir versus placebo because of the limited number of trials. The small-study effect observed could be due to a reporting bias, indicating that negative larger trials have not been published and/or that risk of bias is higher in smaller studies.

Only two trials addressed head-to-head comparisons. Our network meta-analysis revealed no statistically significant differences between drugs. After adjustment for small-study effects, valacyclovir may be superior to acyclovir, and acyclovir and valacyclovir may be superior to famciclovir, but uncertainty is substantial. Pooled analysis of numbers of participants with at least one recurrence virologically confirmed did not provide additional information. Included trials were often older, and confirmation was by viral culture, which lacks sensitivity. In addition, because shedding occurs in the absence of clinical recurrence, this outcome is not clinically relevant. This outcome was often described in the methods sections but was rarely reported in the results sections (Table 3).

The number of recurrences after treatment withdrawal was evaluated in some of the trials. Because these follow-up studies were open-label, results were not analyzed in the review. No conclusion on adverse events was possible because of the poor reporting quality of this outcome and the lack of possible data synthesis.

Overall completeness and applicability of evidence

We were able to analyze only the outcome of at least one recurrence of genital herpes. Survival methods based on time to first recurrence were often used, but the poor and heterogeneous reporting precluded synthesis. As a consequence, there are no data for one of the primary outcomes. Future trial reports should include hazard ratios and associated confidence intervals.

The two methods, ie comparing proportions of participants experiencing at least one recurrence or comparing cumulative recurrence-free survival probabilities over a certain follow-up, are straightforward and unbiased. They are not affected by the potential methodological issue related to the fact that a first recurrence affects the risk of subsequent recurrences or compliance with the allocated treatment. In fact, occurrence of the first recurrence of genital herpes may markedly increase the risk of subsequent recurrence and could lead participants in the placebo arm to initiate active therapy. However, consideration of only first events may lead to underestimation of potential benefits in terms of events prevented by the treatment as seen when the intervention fails to affect the first event but nevertheless influences subsequent events. The total number of recurrences and clinical characteristics of recurrence in terms of duration and severity would be informative. The total number of recurrences was available for 10 trials, but we could not calculate the total number of episodes of recurrence over the follow-up period divided by the total person-time because these latter data were never reported.

We could not analyze the outcome of symptomatic and asymptomatic shedding. This outcome was rarely evaluated, and reporting of data did not allow for synthesis. One objective of suppressive treatment is to lower the risk of transmission. One trial (not included because no minimum number was required for inclusion) demonstrated a decrease in transmission in discordant couples for treated participants compared with those given placebo (Corey 2004b). This situation may be due to decreased viral shedding. This review focused on clinical manifestations of genital herpes. The main result that could have been obtained with this review is that antiviral treatment decreases shedding in patients experiencing at least four recurrences per year. In fact, prevention of transmission is a wider issue that concerns people with or without clinical recurrence. Furthermore, a recent trial demonstrated that shedding is not eradicated by antiviral treatment in symptomatic or asymptomatic patients (Johnston 2012).

One important aim of treatment is to decrease the burden related to clinical symptoms of recurrence. Our included trials provided no data on impact on quality of life, including sexual life, or on burden of this daily long-term treatment (Tran 2012).

Because of the heterogeneity of doses in included trials, we could not determine the best dose-efficacy dosage and the best regimen (total dose given once a day or twice a day). Regarding the dose-effect relationship for valacyclovir, 500 mg may be more effective than 250 mg, but data on the 250 mg dose came from one trial only (Reitano 1998); moreover, because of the heterogeneity between the two trials evaluating the dose of 1000 mg per day, we could not state that 1000 mg is superior to 500 mg. The FDA and European Medicines Agency (EMA) current Summaries of Product Characteristics (SPCs) indicate valacyclovir, 1000 and 500 mg per day, respectively. The FDA SPC states the possibility of a 500-mg daily dose for participants with fewer than nine recurrences per year. For acyclovir and famciclovir, no significant dose-effect relationship was found. The FDA SPC states a 400-mg twice-daily dose for acyclovir, and the FDA and EMA SPCs state a 250-mg twice-daily dose for famciclovir.

These results are in accordance with European guidelines (IUSTI) and current practice. We chose the threshold of four recurrences per year as inclusion criteria. Previous SPCs in some European countries stated that suppressive antiviral treatment was restricted to at least six recurrences per year because of the many trials on this subgroup of participants. Current European (EMA) (http://www.emea.europa.eu/ema/index.jsp?curl=pages/includes/medicines/medicines_landing_page.jsp) (last checked on March 18, 2013) and US (FDA) (http://www.accessdata.fda.gov) (last checked on March 18, 2013) guidelines and recent European guidelines state no minimum number of recurrences required to initiate suppressive treatment. A joint evaluation of burden by physicians and patients is recommended. We excluded studies with no minimum number of recurrences required, which indicates that symptomatic as well as asymptomatic patients could be randomly assigned. We did not find and did not have to exclude any study that listed among its inclusion criteria symptomatic patients with no minimum recurrence required.

Quality of the evidence

The review included 26 trials of 1437 participants randomly assigned to receive acyclovir, 2123 valacyclovir, and 1162 famciclovir. The included trials, especially acyclovir versus placebo trials, were old, with 18 of 26 published before 1999. This fact explained at least in part the poor reporting quality of the trials.

Regarding heterogeneity, large variation in effects was noted for some comparisons (acyclovir vs placebo; valacyclovir vs placebo); others had limited numbers of trials. We did not identify clear reasons for this heterogeneity: as far as we could assess it, clinical and methodological diversity was low to moderate across trials, and we could not perform prespecified subgroup analyses comparing men versus women, HSV-1 versus HSV-2 genital herpes, or treatment regimens according to frequency.

Regarding indirectness, we found no evidence of inconsistency between direct and indirect effect estimates. However, the test of consistency based on inconsistency factors has typically low power, and statistically nonsignificant inconsistency does not necessarily imply clinical consistency. We tried to assess transitivity by comparing trial and participant characteristics across comparisons (Table 2) and found that trials were similar across comparisons. Of note, publication year (as a proxy for time of trial conduct) varied across comparisons. The variations correspond logically to the sequence of development of these drugs. However, no clear dependence of treatment effects on publication year was noted and older trials did not systematically show higher risk of bias as compared with more recent trials.

Regarding risk of bias, no trial was determined to be at low risk of bias. The level of risk of bias was considered high for half of the trials and unclear for the other half. Risk of bias for generation of sequence generation and allocation concealment was most frequently unclear. Risk of bias for blinding was most frequently low. Risk of bias was considered high for almost one-third of the trials because of incomplete outcome data resulting from high or unbalanced rates of withdrawn participants. For most trials, analyses were not performed according to intention-to-treat principles, and the number of analyzed participants differed from the number of randomly assigned participants. For the main analysis, given that the mean number of recurrences among included participants was 11 per year, that the mean duration of treatment was about six months, and that the risk of at least one recurrence was greater than 40% for the treated group and greater than 80% for the nontreated group, we assumed that any participant with missing outcome data experienced at least one recurrent episode, whatever the group. Because no protocol was available for any of the included trials, the risk of bias for selective outcome reporting was considered unclear for most. Risk of bias for selective outcome reporting was considered high for about one-fifth of the trials because some outcomes stated in the methods sections were not reported in the results sections. In addition, in the four trials included but not analyzed because the main outcome was not available (even after study authors were contacted), investigators likely measured the recurrence outcome but did not report it. One of these four trials reported the total number of recurrences; others focused on shedding but aimed to differentiate symptomatic and asymptomatic shedding; thus data on numbers of participants with recurrence were logically available.

Lastly, cross-over trials were not reported appropriately, in that the data required to incorporate paired analyses in our meta-analyses were not available. As a consequence, we used a simple approach to incorporating these trials in syntheses, which gives rise to a unit-of-analysis error. However, we found that effect estimates were larger in cross-over trials than in parallel-group trials so we did not combine the two types of trial designs, and we based our conclusions on parallel-group trials, ie the GRADE assessments and Summary of Findings table data are based on effect estimates from parallel group studies.

Regarding publication bias and related small-study effects, we identified a small-study effect in the comparison of acyclovir and placebo; it was not possible to assess the effect for other comparisons. Treatment effect estimates of the three drugs against placebo were reduced after adjustment for small-study effects.

Overall, we are confident that the three antiviral drugs are superior to placebo, but, in light of heterogeneity and small-study effects, true effects may be substantially different from effect estimates. Moreover, we are moderately confident that the three drugs are not different in terms of indirectness (few head-to-head trials were identified) and uncertainty (confidence intervals for pairwise comparisons between the three antiviral drugs are wide).

Potential biases in the review process

We performed a wide search for trials, including contacting pharmaceutical companies and trial authors, and searched the US FDA database and abstract proceedings of seven congresses up to a maximum of 10 years. The probability that we missed a trial is thus low. However, at least for acyclovir trials, the funnel plots highlight a small-study effect, that could be due to publication bias. In fact, many trials were performed in the 1980s and 1990s, before the International Committee of Medical Journal Editors issued the requirement of trial registration for publication. The possibility that we may have missed a trial is thus not excluded.

Agreements and disagreements with other studies or reviews

A previous systematic review and meta-analysis of the same topic included 14 trials and also found that acyclovir, famciclovir, and valacyclovir were all superior to placebo for the outcome of risk of at least one clinical recurrence of genital herpes (Lebrun-Vignes 2007).

Authors' conclusions

Implications for practice

Owing to risk of bias and inconsistency, there is a low quality evidence that suppressive antiviral therapy with acyclovir, valacyclovir or famciclovir for participants experiencing at least 4 recurrences per year of genital herpes decreases the number of participants with at least one recurrence as compared with placebo. Network meta-analysis could not reveal a superiority of one drug over another. The dose used in the included and analyzed studies was from 400 to 800 mg per day for acyclovir, 250 to 1000 mg per day for valacyclovir and 125 to 750 mg per day for famciclovir. There was no clear evidence of a dose-effect relationship for any drug. The duration of suppressive treatment was from 2 to 12 months.

Implications for research

Inclusion criteria in future trials should be decided according to current practice based on the impact of genital herpes recurrences on quality of life rather than on the minimum number of recurrences. Comparative trials between acyclovir or valacyclovir or famciclovir versus placebo are no longer suitable because the efficacy of these treatments has been demonstrated. A large pragmatic trial of a direct comparison between these drugs or between new antiviral drugs and these drugs is needed. The main outcome should be a patient-reported outcome including burden of treatment and impact on quality of life, notably sexual life. Secondary outcomes should include safety, time to first recurrence, number of recurrences, and severity (number of days, severity of pain) of recurrences. The duration of treatment could also be an outcome according to patient wishes.

Acknowledgements

Pr Corey and Pr Wald for their help on trials and data research and for their comments.

Data and analyses

Download statistical data

Comparison 1. Acyclovir vs placebo
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Particpants with at least 1 clinical recurrence14 Risk Ratio (M-H, Random, 95% CI)Subtotals only
1.1 Parallel group design92049Risk Ratio (M-H, Random, 95% CI)0.48 [0.39, 0.58]
1.2 Crossover design5500Risk Ratio (M-H, Random, 95% CI)0.35 [0.29, 0.42]
2 Participants with at least 1 clinical recurrence (according to daily dose)9 Risk Ratio (M-H, Random, 95% CI)Subtotals only
2.1 Acyclovir 400 mg/d1102Risk Ratio (M-H, Random, 95% CI)0.39 [0.27, 0.56]
2.2 Acyclovir 600 mg/d3140Risk Ratio (M-H, Random, 95% CI)0.38 [0.13, 1.06]
2.3 Acyclovir 800 mg/d51756Risk Ratio (M-H, Random, 95% CI)0.52 [0.43, 0.64]
2.4 Acyclovir 1000 mg/d1101Risk Ratio (M-H, Random, 95% CI)0.40 [0.28, 0.57]
3 Participants with at least 1 clinical recurrence (complete cases)14 Risk Ratio (M-H, Random, 95% CI)Subtotals only
3.1 Parallel-group design91797Risk Ratio (M-H, Random, 95% CI)0.46 [0.38, 0.56]
3.2 Cross-over design5406Risk Ratio (M-H, Random, 95% CI)0.20 [0.14, 0.28]
4 Particpants with at least 1 virologically confirmed recurrence3139Risk Ratio (M-H, Random, 95% CI)0.08 [0.03, 0.22]
4.1 Parallel-group design3139Risk Ratio (M-H, Random, 95% CI)0.08 [0.03, 0.22]
Analysis 1.1.

Comparison 1 Acyclovir vs placebo, Outcome 1 Particpants with at least 1 clinical recurrence.

Analysis 1.2.

Comparison 1 Acyclovir vs placebo, Outcome 2 Participants with at least 1 clinical recurrence (according to daily dose).

Analysis 1.3.

Comparison 1 Acyclovir vs placebo, Outcome 3 Participants with at least 1 clinical recurrence (complete cases).

Analysis 1.4.

Comparison 1 Acyclovir vs placebo, Outcome 4 Particpants with at least 1 virologically confirmed recurrence.

Comparison 2. Valacyclovir vs placebo
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Participants with at least 1 clinical recurrence4 Risk Ratio (M-H, Random, 95% CI)Subtotals only
1.1 Parallel-group design41788Risk Ratio (M-H, Random, 95% CI)0.41 [0.24, 0.69]
2 Participants with at least 1 clinical recurrence (according to daily dose)4 Risk Ratio (M-H, Random, 95% CI)Subtotals only
2.1 Valacyclovir 250 mg/d1403Risk Ratio (M-H, Random, 95% CI)0.81 [0.73, 0.90]
2.2 Valacyclovir 500 mg/d31101Risk Ratio (M-H, Random, 95% CI)0.45 [0.30, 0.67]
2.3 Valacyclovir 1000 mg/d2552Risk Ratio (M-H, Random, 95% CI)0.39 [0.13, 1.18]
3 Participants with at least 1 clinical recurrence (complete cases)41787Risk Ratio (M-H, Random, 95% CI)0.41 [0.25, 0.69]
3.1 Parallel-group design41787Risk Ratio (M-H, Random, 95% CI)0.41 [0.25, 0.69]
Analysis 2.1.

Comparison 2 Valacyclovir vs placebo, Outcome 1 Participants with at least 1 clinical recurrence.

Analysis 2.2.

Comparison 2 Valacyclovir vs placebo, Outcome 2 Participants with at least 1 clinical recurrence (according to daily dose).

Analysis 2.3.

Comparison 2 Valacyclovir vs placebo, Outcome 3 Participants with at least 1 clinical recurrence (complete cases).

Comparison 3. Valacyclovir vs No treatment
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Participants with at least 1 clinical recurrence2 Risk Ratio (M-H, Random, 95% CI)Subtotals only
1.1 Parallel-group design1237Risk Ratio (M-H, Random, 95% CI)0.46 [0.37, 0.56]
1.2 Cross-over design1450Risk Ratio (M-H, Random, 95% CI)0.57 [0.50, 0.66]
2 Participants with at least 1 clinical recurrence (complete cases)2641Risk Ratio (M-H, Random, 95% CI)0.49 [0.43, 0.56]
2.1 Parallel-group design1237Risk Ratio (M-H, Random, 95% CI)0.46 [0.37, 0.56]
2.2 Cross-over design1404Risk Ratio (M-H, Random, 95% CI)0.51 [0.43, 0.61]
Analysis 3.1.

Comparison 3 Valacyclovir vs No treatment, Outcome 1 Participants with at least 1 clinical recurrence.

Analysis 3.2.

Comparison 3 Valacyclovir vs No treatment, Outcome 2 Participants with at least 1 clinical recurrence (complete cases).

Comparison 4. Famciclovir vs placebo
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Participants with at least 1 clinical recurrence2 Risk Ratio (M-H, Random, 95% CI)Subtotals only
1.1 Parallel-group design2832Risk Ratio (M-H, Random, 95% CI)0.57 [0.50, 0.64]
2 Particpants with at least 1 clinical recurrence (according to daily dose)2 Risk Ratio (M-H, Random, 95% CI)Subtotals only
2.1 Famciclovir 125 mg/d1124Risk Ratio (M-H, Random, 95% CI)0.79 [0.55, 1.14]
2.2 Famciclovir 250 mg/d1190Risk Ratio (M-H, Random, 95% CI)0.67 [0.48, 0.92]
2.3 Famciclovir 325 mg/d1226Risk Ratio (M-H, Random, 95% CI)0.61 [0.50, 0.75]
2.4 Famciclovir 500 mg/d2420Risk Ratio (M-H, Random, 95% CI)0.53 [0.44, 0.64]
2.5 Famciclovir 750 mg/d1228Risk Ratio (M-H, Random, 95% CI)0.52 [0.41, 0.65]
3 Participants with at least 1 clinical recurrence (complete cases)2 Risk Ratio (M-H, Random, 95% CI)Subtotals only
3.1 Parallel-group design2711Risk Ratio (M-H, Random, 95% CI)0.45 [0.23, 0.86]
4 Participants with at least 1 virologically confirmed recurrence2555Risk Ratio (M-H, Random, 95% CI)0.41 [0.25, 0.65]
4.1 Parallel-group design2555Risk Ratio (M-H, Random, 95% CI)0.41 [0.25, 0.65]
Analysis 4.1.

Comparison 4 Famciclovir vs placebo, Outcome 1 Participants with at least 1 clinical recurrence.

Analysis 4.2.

Comparison 4 Famciclovir vs placebo, Outcome 2 Particpants with at least 1 clinical recurrence (according to daily dose).

Analysis 4.3.

Comparison 4 Famciclovir vs placebo, Outcome 3 Participants with at least 1 clinical recurrence (complete cases).

Analysis 4.4.

Comparison 4 Famciclovir vs placebo, Outcome 4 Participants with at least 1 virologically confirmed recurrence.

Comparison 5. Valacyclovir vs acyclovir
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Participants with at least 1 clinical recurrence1 Risk Ratio (M-H, Random, 95% CI)Subtotals only
1.1 Parallel-study design11345Risk Ratio (M-H, Random, 95% CI)1.16 [1.01, 1.34]
2 Participants with at least 1 clinical recurrence (complete cases)1 Risk Ratio (M-H, Random, 95% CI)Subtotals only
2.1 Parallel-study design11345Risk Ratio (M-H, Random, 95% CI)1.16 [1.01, 1.34]
Analysis 5.1.

Comparison 5 Valacyclovir vs acyclovir, Outcome 1 Participants with at least 1 clinical recurrence.

Analysis 5.2.

Comparison 5 Valacyclovir vs acyclovir, Outcome 2 Participants with at least 1 clinical recurrence (complete cases).

Comparison 6. Famciclovir vs valacyclovir
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Participants with at least 1 clinical recurrence1 Risk Ratio (M-H, Random, 95% CI)Subtotals only
1.1 Parallel-study design1320Risk Ratio (M-H, Random, 95% CI)1.18 [0.86, 1.63]
2 Participants with at least 1 clinical recurrence (complete cases)1 Risk Ratio (M-H, Random, 95% CI)Subtotals only
2.1 Parallel-study design1315Risk Ratio (M-H, Random, 95% CI)1.21 [0.87, 1.69]
3 Participants with at least 1 virologically confirmed recurrence1 Risk Ratio (M-H, Random, 95% CI)Subtotals only
3.1 Parallel-study design1315Risk Ratio (M-H, Random, 95% CI)2.24 [1.05, 4.76]
Analysis 6.1.

Comparison 6 Famciclovir vs valacyclovir, Outcome 1 Participants with at least 1 clinical recurrence.

Analysis 6.2.

Comparison 6 Famciclovir vs valacyclovir, Outcome 2 Participants with at least 1 clinical recurrence (complete cases).

Analysis 6.3.

Comparison 6 Famciclovir vs valacyclovir, Outcome 3 Participants with at least 1 virologically confirmed recurrence.

Appendices

Appendix 1. MEDLINE search strategy

MEDLINE search strategy using the HIV/AIDS group filter

("Acyclovir"[Mesh:NoExp] OR "valacyclovir "[Substance Name] OR "famciclovir "[Substance Name] OR "Antiviral Agents"[Mesh:NoExp] OR "Antiviral Agents "[Pharmacological Action] OR suppressive treatment[Title/Abstract] OR suppressive therapy[Title/Abstract] OR suppressive agent*[Title/Abstract] OR suppressive drug*[Title/Abstract] OR antiviral drug*[Title/Abstract] OR antiviral therapy[Title/Abstract] OR antiviral treatment[Title/Abstract] OR antiviral agent*[Title/Abstract]) AND ("Herpes Genitalis"[Mesh] OR "Herpes Simplex"[Mesh:NoExp] OR hsv2[Title/Abstract] OR hsv 2[Title/Abstract] OR genital herpes[Title/Abstract] OR (("Herpes Simplex"[Mesh:NoExp] OR "Simplexvirus"[Mesh:NoExp]) AND genital[Title/Abstract])) AND ((randomized controlled trial [pt] OR controlled clinical trial [pt] OR randomized controlled trials [mh] OR random allocation [mh] OR double-blind method [mh] OR single-blind method [mh] OR clinical trial [pt] OR clinical trials [mh] OR ("clinical trial" [tw]) OR ((singl* [tw] OR doubl* [tw] OR trebl* [tw] OR tripl* [tw]) AND (mask* [tw] OR blind* [tw])) OR (placebos [mh] OR placebo* [tw] OR random* [tw] OR research design [mh:noexp] OR (comparative trial) OR (comparative studies) OR (evaluation studies) OR (evaluation trial) OR follow-up studies [mh] OR prospective studies [mh] OR control*[tw] OR prospectiv* [tw] OR volunteer* [tw]) NOT (animals [mh] NOT human [mh]))

 

MEDLINE search strategy using the high sensitive Cochrane filter

("Acyclovir"[Mesh:NoExp] OR "valacyclovir "[Substance Name] OR "famciclovir "[Substance Name] OR "Antiviral Agents"[Mesh:NoExp] OR "Antiviral Agents "[Pharmacological Action] OR suppressive treatment[Title/Abstract] OR suppressive therapy[Title/Abstract] OR suppressive agent*[Title/Abstract] OR suppressive drug*[Title/Abstract] OR antiviral drug*[Title/Abstract] OR antiviral therapy[Title/Abstract] OR antiviral treatment[Title/Abstract] OR antiviral agent*[Title/Abstract]) AND ("Herpes Genitalis"[Mesh] OR "Herpes Simplex"[Mesh:NoExp] OR hsv2[Title/Abstract] OR hsv 2[Title/Abstract] OR genital herpes[Title/Abstract] OR (("Herpes Simplex"[Mesh:NoExp] OR "Simplexvirus"[Mesh:NoExp]) AND genital[Title/Abstract])) AND ((randomized controlled trial [pt] OR controlled clinical trial [pt] OR randomized [tiab]  OR placebo [tiab]  OR drug therapy [sh]  OR randomly [tiab]  OR trial [tiab]  OR groups [tiab]) NOT (animals [mh] NOT humans [mh]))

Appendix 2. EMBASE search strategy

((suppressive OR antiviral) NEAR/1 (agent* OR drug* OR therapy OR treatment)):ab,ti OR 'antivirus-agent':de OR acyclovir:de OR 'aciclovir':de OR 'valaciclovir':de OR valacyclovir:de OR 'famciclovir':de OR famcyclovir:de OR acyclovir:ab,ti OR aciclovir:ab,ti OR valaciclovir:ab,ti OR valacyclovir:ab,ti OR famciclovir:ab,ti OR famcyclovir:ab,ti AND (herpes:de AND simplex:de AND virus:de AND genital:ab,ti OR 'herpes simplex virus 2':de OR 'genital herpes':de OR (genital NEAR/1 herpes):ab,ti) AND (random* OR factorial* OR crossover* OR (cross AND over*) OR placebo* OR doubl* NEAR/1 blind* OR singl* NEAR/1 blind* OR assign* OR allocat* OR volunteer* OR 'crossover procedure'/exp OR 'double blind procedure'/exp OR 'randomized controlled trial'/exp OR 'single blind procedure'/exp) AND [embase]/lim

Appendix 3. CENTRAL search strategy

((genital herpes):ti,ab,kw AND (hsv 2):ti,ab,kw OR (Herpes Genitalis):kw OR (hsv2):ti,ab,kw OR (Herpesvirus 2, Human):kw OR ((Herpes Simplex):kw OR (Simplexvirus):kw) AND (genital):ti,ab,kw ) AND ((suppressive agent*):ti,ab,kw OR (antiviral agent*):ti,ab,kw OR (antiviral treatment):ti,ab,kw OR (antiviral therapy):ti,ab,kw OR (antiviral drug*):ti,ab,kw OR (suppressive drug*):ti,ab,kw OR (suppressive therapy):ti,ab,kw OR (suppressive treatment):ti,ab,kw OR (antiviral agents):kw OR (famciclovir):ti,ab,kw OR (famcyclovir):ti,ab,kw OR (acyclovir):ti,ab,kw OR (aciclovir):ti,ab,kw OR (valacyclovir):ti,ab,kw OR (valaciclovir):ti,ab,kw)

Appendix 4. LILACS search strategy

genital herpes [Words] or Herpes Genitalis [DeCS Category]

Contributions of authors

L. Le Cleach and L Trinquart contributed equally to this study.

O Chosidow and B Lebrun-Vignes have performed a previous work that is the foundation of the current work.

L Le Cleach, L Trinquart, O Chosidow and P Ravaud have conceived of and designed the review and the search strategy and interpreted the data.

L Le Cleach and L Trinquart undertook searches including unpublished studies, and entered data.

L Le Cleach, G Do and L Trinquart screened retrieved papers against eligibility criteria.

L Le Cleach, A Maruani and L Trinquart evaluated the risk of bias and extracted data from papers.

L Trinquart and L Le Cleach analyzed the data.

L Le Cleach and L. Trinquart wrote the review.

Declarations of interest

L Le Cleach, L. Trinquart, Giao Do, B Lebrun-Vignes and Philippe Ravaud have no conflicts of interest to declare

O Chosidow was a consultant for GSK.

Sources of support

Internal sources

  • No sources of support supplied

External sources

  • Association Recommandations en Dermatologie (aRED), a group of the French Society of Dermatology (SFD), France.

    a grant

  • French Ministry of Health, France.

    Grant support was from the Programme Hospitalier de Recherche Clinique (AOM 11151)

  • The funding agencies have no role in the design or conduct of the study; collection, management, analysis, or interpretation of the data; or preparation and review of the manuscript,, Other.

Differences between protocol and review

  • In the protocol, we stated that we would discuss the eligibility of quasi-randomized trials based on the number of included randomized trials. In fact, we did not identify any eligible quasi-randomized trial from our searches.

  • In the review, we added one secondary outcome (proportion of participants with at least one virologically confirmed recurrent episode of genital herpes during the treatment period), as, during the review process, we found it relevant and it was reported in several trials.

  • We could not analyze the incidence rates of recurrences as planned. We aimed to count the total number of recurrent episodes over the follow-up period and to divide it by the total amount of person-time.

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Blom 1986

Methods

Prospective, double-blind, randomised, cross-over trial

5 centres in Sweden

Period of inclusion not stated

Participants

Inclusion criteria

  • Recurrent genital herpes with a virologically proven HSV2 infection (method for virological confirmation not stated)

Exclusion criteria

  • Pregnancy

Baseline data : not on randomised: (39); on analyzed participants (33)

  • Mean age (extremes), years: 32 (20-63)

  • Male/female: 22/11

  • Duration of condition, (years): not stated

  • Mean number of recurrences per year (extremes): 12.5 (8-21)

  • Number of days between last recurrence and initiation of treatment : 0 (at onset of relapse participants were given the first treatment)

Withdrawal: acyclovir 200 x 4/d ; placebo

  • Adverse events: 3 (all during ACV treatment): stop at days 53, 15, and 46

  • Bad compliance: 3

Interventions

Duration of periods of treatment: until first recurrence or maximum 12 weeks

Duration of wash-out between periods of treatment: not stated

Intervention 1: oral acyclovir 200 mg x 4/d

Intervention 2: placebo

Co-intervention: none

Outcomes

Primary outcome: not stated

Definition of recurrence: recurrence observed by a physician or reported by a participant

Outcomes:

  • Number of participants with active lesions

  • Median time to recurrences

  • Adverse events

Method of recurrences assessment:

  • Scheduled visit each 4 weeks

  • Unplanned visit motivated by recurrences

  • Participant diary: number of tablets taken, any symptoms or signs of recurrent attack

  • Swab for HSV viral culture during recurrences

NotesNo funding reported; however belongs to a group of 5 cross-over trials (Blom 1986; Halsos 1985; Kinghorn 1985; Kroon 1989; Thin 1985)having very similar design of which 4 were declared funded by Wellcome
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear risk

Quote: "randomised"

Comment: method used to generate the allocation sequence was not reported

Allocation concealment (selection bias)Unclear riskComment: no description of the method used to guarantee allocation concealment
Blinding of participants and personnel (performance bias)
All outcomes
Low risk

Quote: "double blind"

Comment: no precision but as placebo controlled trial probably done

Blinding of outcome assessment (detection bias)
All outcomes
Low risk

Quote: "double blind"

Comment: no precision but as placebo controlled trial probably done

Incomplete outcome data (attrition bias)
All outcomes
Low risk

Number of randomly assigned participants: 39

Number of analyzed participants for time to first recurrence and rate of recurrence: 33

  • Adverse events: 3 (all during ACV treatment) stop at day 53, 15 and 46

  • Bad compliance: 3

Selective reporting (reporting bias)High riskComment: no protocol available, no results per period

Diaz-Mitoma 1998

Methods

Prospective, double-blind, randomized, parallel-group trial

30 centers in Canada and Europe (Belgium, France, Iceland, Sweden, United Kingdom)

Period of inclusion not stated

Participants

Inclusion criteria

  • Age ≥ 18 years

  • Recurrent genital herpes confirmed by herpes serology OR HSV viral culture

  • At least 6 recurrences of genital herpes in the past year or in any 12-month period in the 2 years before trial entry while not receiving antiviral treatment

  • 14 days of washout of previous antiviral treatment

Exclusion criteria

  • Pregnancy/lactation

  • Immunosuppression

Baseline data

Randomly assigned: famciclovir 125 mg x 3 (112), famciclovir 250 x 2 (117), famciclovir 250 x 3 (114), placebo (114)

  • Mean age (extremes), years: 35 (19-67); 38 (20-76); 37 (20-66); 37 (20-68)

  • Male/ female: 52/60; 64/52; 59/64; 48/66

  • Duration of condition, (years): 6.7 (1-25); 7.2 (1-33); 7 (1-38); 6.7 (0-25)

Withdrawal: famciclovir 125 mg x 3 (48), famciclovir 250 x 2 (42), famciclovir 250 x 3 (40), placebo (88)

  • Lack of efficacy: 21; 17; 18; 63

  • Adverse events: 6; 5; 4; 6

  • Other reasons: 21; 20; 18; 19

Interventions

Duration of treatment: 52 weeks

Intervention 1: oral famciclovir 125 mg x 3/d

Intervention 2: oral famciclovir 250 mg x 2/d

Intervention 3: oral famciclovir 250 mg x 3/d

Intervention 4: placebo

Co-intervention: none

Outcomes

In methods section

Primary outcome: time to first recurrence and % of participants free of recurrence and % of participants free of recurrence at least 6 months after start of trial

Definition of recurrence: viral confirmation needed

Secondary outcomes:

  • % of participants free of recurrence at least 12 months after start of trial

  • Median number of recurrences per year

  • Adverse events

Method of recurrences assessment:

  • Scheduled visit each 28 days

  • Unplanned visit motivated by recurrences within 24 hours following onset of recurrence

  • Patient diary: self assessment episodes

  • Swab for HSV viral culture during recurrences

Notes

Trial funded by SmithKline Beecham Pharmaceuticals

2 of 5 authors were declared as SmithKline Beecham Pharmaceuticals

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk

Quote: "a unique patient number at each centre according to computer-generated randomisation code"

Comment: probably done

Allocation concealment (selection bias)Low risk

Quote: "the trial medication was blister packed and packs were labelled with the patient number"

Comment: probably done

Blinding of participants and personnel (performance bias)
All outcomes
Low risk

Quote: "the patient, the investigator and the sponsor personnel directly involved in monitoring the trial or reviewing data had no knowledge of what treatment had been allocated until the code was unblinded and the data were analyzed. participants took 7 tablets daily during each regimen. Tablets containing placebo were identical to those containing each dosage of famciclovir. There were no reports of compromised blinding by vision or taste"

Comment: probably done

Blinding of outcome assessment (detection bias)
All outcomes
Low risk

Quote: "the investigator and the sponsor personnel directly involved in monitoring the trial or reviewing data had no knowledge of what treatment had been allocated until the code was unblinded and the data were analyzed".

Comment: probably done

Incomplete outcome data (attrition bias)
All outcomes
High risk

Number of randomly assigned participants: famciclovir 125 mg x 3 (112), famciclovir 250 x 2 (117), famciclovir 250 x 3 (114), placebo (114)

Number of analyzed participants

  • Time to first recurrence: 112, 116, 113, 114 (all participants who received at least one dose)

  • Rates of recurrence at 12 months: 77, 90, 81, 88 (population known to have experienced at least 1 recurrence or to have been recurrence free during the entire reference time interval)

Missing data management for time to first recurrence: no information

Comment: large quantity of missing data (for rate of recurrence at 12 months)

Selective reporting (reporting bias)Unclear riskComment: protocol not available

Douglas 1984

Methods

Prospective, double-blind, randomized, parallel-group trial

One center in USA

From October 1981 to January 1983

Participants

Inclusion criteria

  • Recurrent genital herpes confirmed by HSV2 positive viral culture

  • At least 6 recurrences of genital herpes in the previous 12 month

Exclusion criteria

  • Pregnancy

  • Immunosuppression

  • Any immunosuppressive or antiviral therapy

  • Clinical recurrence at the time of inclusion

Baseline data:

  • Randomly assigned: acyclovir 200 mg x 5 (51), acyclovir 200 mg x 2 (52), placebo (50 (26 5/d and 24 2/d))

  • Mean age (SD), years: 30.6(6.1); 32.5 (6.9); 32.4 (7.9)

  • Male/ female: 25/26; 26/26; 24/26

  • Duration of condition (median months) : 33; 31;27

  • Mean number of recurrences past year (SD): 13.3(7.9), 12.1 (4.2); 13.4 (6)

  • Number of days between last recurrence and initiation of treatment (SD): 16.1 (11.6); 15.3(9); 18.9(15.1)

Withdrawal: acyclovir 200 mg x 5 (6), acyclovir 200 mg x 2 (1), placebo (3)

  • Adverse event: 1; 0; 0

  • Change of residence: 4 in all

  • Administrative reasons: 5 in all

Interventions

Durationof treatment: 4 months

Intervention 1: oral acyclovir 200 mg x 5/d

Intervention 2: oral acyclovir 200 mg x 2/d

Intervention 3: placebo 2/d or 5/d

Co-intervention:none

Outcomes

In results section

Primary outcome: not specified

Definition of recurrence: recurrence observed by a physician or reported by a participant

Outcomes

  • Median time to first recurrence

  • % of patient with at least a recurrence

  • Mean number of recurrences per month

  • Mean number of prodromal episodes

  • Median time to complete healing of lesion and cessation of local pain

  • Number of positive viral cultures

  • Compliance

  • Mean acyclovir plasma level

  • Adverse events

Methods of recurrence assessment

  • Scheduled visit monthly

  • Unplanned visit motivated by recurrences within 48 hours following onset of recurrence $

  • No participant diary

  • Swab of the lesion for HSV-2 viral culture during recurrences

NotesQuote: "supported by grants from the Burroughs Wellcome Company"
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear risk

Quote:"randomisation codes established in blocks of six"

Comment: process of selecting blocks was not specified

Allocation concealment (selection bias)Unclear riskComment: no description of the method used to guarantee allocation concealment
Blinding of participants and personnel (performance bias)
All outcomes
Low risk

Quote: "double-blind and drug and placebo were identical in appearance"

Comment: no precision but as placebo controlled trial probably done

Blinding of outcome assessment (detection bias)
All outcomes
Low risk

Quote: "all efficacy and toxicity measures were coded without knowledge of group assignment"

Comment: no precision but as placebo controlled trial probably done

Incomplete outcome data (attrition bias)
All outcomes
Low risk

Randomly assigned: acyclovir 200 mg x 5 (51), acyclovir 200 mg x 2 (52), placebo (50)

Analyzed for

  • Time to first recurrence: 51, 52, 50

  • Rate of recurrence: 45, 51, 47

Management of missing data: participants with missing values for particular variables were excluded from analyses that involved those variables

Selective reporting (reporting bias)Unclear riskComment: no protocol available and outcomes not clearly announced in the methods section; rate of clinical recurrence and time to first recurrence fully reported but unclear whether rate of recurrence corresponded to recurrence virologically confirmed or only clinically confirmed. No primary outcome stated

Fife 2006

Methods

Prospective, double-blind, randomized, parallel-group trial

27 centers in USA

From June 18, 2004 to December 17, 2004

Participants

Inclusion criteria

  • Age ≥ 18 years

  • Recurrent genital herpes confirmed by a positive HSV-2 serology

  • At least 6 recurrences of genital herpes per year

Exclusion criteria

  • Pregnancy

  • Immunosuppression

  • Impaired renal or hepatic function

Baseline data:

  • Randomly assigned: valacyclovir (109); placebo (43)

  • Mean age (extremes), years : 39.8 (18-66); 40.7 (19-61)

  • Male/female: 39/70; 12/30

Withdrawal:valacyclovir 500mg x2 (15); placebo (3)

  • Adverse events: 1; 0

  • Decision to withdraw: 8; 3

  • Lost to follow-up: 4; 0

  • Protocol violation: 2; 0

Interventions

Duration of treatment: 60 days

Intervention 1: oral valacyclovir 500 mg x 2/d

Intervention 2: placebo 2/d

Co-intervention: valacyclovir 500 mg x 2/d during 3 days in case of recurrence

Outcomes

Primary outcome

% of days with total (clinical and subclinical) HSV-2 shedding as determined by type-specific PCR assay for HSV-2

Definition of clinical recurrence: unclear wether only physician reported or also participant self reported

Secondary outcomes

  • % of clinical days with HSV-2 shedding

  • % of subclinical days with HSV-2 shedding

  • Average log HSV-2 DNA copy number

  • % of participants with 1 or more days of shedding

  • Time to first recurrence

Methods of recurrence assessment

Scheduled visit each 15 days ± 5 days

Unplanned visit motivated by recurrences

Shedding: daily genital + anal + rectal swab for PCR; no extra sampling on recurrence site

Notes

3 of 7 authors were declared as GlaxoSmithKline employees

Quote:"financial support for this trial was provided by GlaxoSmithKline..."

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear risk

Quote: "central randomisation"

Comment: process for generating the allocation sequence was not reported

Allocation concealment (selection bias)Low risk

Quote: "central randomizations and film coated tablets"

Comment: probably done

Blinding of participants and personnel (performance bias)
All outcomes
Low risk

Quote: "double blind"

Comment: no precision but as placebo-controlled trial probably done

Blinding of outcome assessment (detection bias)
All outcomes
Low risk

Quote: "double blind"

Comment: no precision but as placebo-controlled trial probably done

Incomplete outcome data (attrition bias)
All outcomes
Low risk

Randomly assigned: valacyclovir (109); placebo (43)

Analyzed for

  • Time to first recurrence and rate of recurrence: 109/42

  • Shedding: 105/40

Management of missing data: "If the PCR data were missing for both sites, the patient shedding status was missing for that day and was excluded from both the numerator and the denominator of calculation to determine the percentage of days with shedding ...."

Selective reporting (reporting bias)Unclear riskComment: protocol not available

Halsos 1985

Methods

Prospective, double-blind, randomized, cross-over trial

2 centers in Norway and Finland

Period of inclusion not stated

Participants

Inclusion criteria

  • Age ≥ 18 years

  • Men

  • Recurrent genital herpes confirmed by positive viral culture

  • At least 8 recurrences of genital herpes per year

Exclusion criteria

  • Other antiviral therapy

Baseline data

Randomly assigned: 31

  • Mean age, years: 31,1

  • Male/female: 31/0

  • Duration of condition, (years): 2.3

  • Mean number of recurrences per year: 11

  • Number of days between last recurrence and onset of treatment: 8.1

Withdrawal: acyclovir 200x4 /d ; placebo

  • 1 at day 9 (reason not stated)

Interventions

Duration of periods of treatment: 12 weeks

Duration of wash-out between periods of treatment: not stated

Intervention 1: oral acyclovir 200 mg x 4 /d

Intervention 2: placebo

Co-intervention: none

Outcomes

Primary outcome: not stated

Definition of recurrence: recurrence observed by a physician or reported by a participant defined as a lesion developing normally

Outcomes:

  • Number of participants experiencing recurrences or abortive episodes

  • Average duration of episodes

  • Adverse events

Method of recurrences assessment:

  • Scheduled visit each 4 weeks

  • Unplanned visit motivated by recurrences when possible

  • Participant diary: compliance; signs and symptoms of recurrence

Notes

Welcome Research laboratories processed and analyzed data

2 study authors are respectively employees of the Welcome foundation, Oslo, Norway, and Welcome Research Laboratories, Beckenham, Kent, England

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear risk

Quote: "randomised"

Comment: method used to generate the allocation sequence was not reported

Allocation concealment (selection bias)Unclear riskComment: no description of the method used to guarantee allocation concealment
Blinding of participants and personnel (performance bias)
All outcomes
Low risk

Quote: "double blind"

Comment: no precision but as placebo-controlled trial probably done

Blinding of outcome assessment (detection bias)
All outcomes
Low risk

Quote: "double blind"

Comment: no precision but as placebo-controlled trial probably done

Incomplete outcome data (attrition bias)
All outcomes
Low risk

Number of randomly assigned participants: 31

Number of analyzed participants for time to first recurrence: 31; rate of recurrence: 30

Participants withdrawn: 1 at day 9 receiving PBO, reason ?

Selective reporting (reporting bias)High riskComment: no results per period, no primary outcome stated, no protocol available

HS240017 2005

Methods

Prospective, double-blind, randomized, parallel-group trial

5 centers in USA

From April 30, 2001, to 03 March 3, 2002

Participants

Inclusion criteria

  • Age ≥ 18 years

  • Recurrent genital herpes

  • 2 to 9 recurrences of genital herpes in the previous year or 1 during the previous 6 months

  • 30-day washout period required for antiviral treatment

Exclusion criteria

  • Pregnancy/lactation

  • Positive human immunodeficiency virus serology at screening

  • Impaired renal function

  • History of liver disease

Baseline data:

  • Randomly assigned: valacyclovir (42); placebo (44)

  • Mean age, years (SD): 40.1 (10.6); 37.9 (10.3)

  • Male/female: 13/29; 11/33

Withdrawal :valacyclovir (7); placebo (9)

Interventions

Duration of treatment: 60 days

Intervention 1: oral valacyclovir 500 mg x 1/d

Intervention 2: placebo

Co-intervention: valacyclovir 500 mg x 2/d during 5 days in case of recurrence

Outcomes

Primary outcome: % of subclinical days with HSV viral shedding as evaluated by PCR

Definition of recurrence: recurrence observed by a physician OR reported by participant self assessment

Secondary outcomes

  • Duration of recurrences

  • Time to first shedding day

  • Duration of clinical shedding

  • Duration of subclinical shedding

  • Total duration of shedding

  • % of days with clinical shedding

  • % of days with shedding

Methods of recurrence assessment

Shedding: daily genital and anal/rectal swab for HSV PCR

Notes 
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear risk

Quote: "randomised"

Comment: method used to generate the allocation sequence was not reported

Allocation concealment (selection bias)Unclear riskComment: no description of the method used to guarantee allocation concealment
Blinding of participants and personnel (performance bias)
All outcomes
Low risk

Quote: "double blind...2 identically appearing oral treatment"

Comment: probably done

Blinding of outcome assessment (detection bias)
All outcomes
Low risk

Quote: "double blind"

Comment: no precision but as placebo-controlled trial probably done

Incomplete outcome data (attrition bias)
All outcomes
Unclear risk

Randomly assigned: valacyclovir (42); placebo (44)

Analyzed for shedding: 35/37

Withdrawal: valacyclovir (7); placebo (9)

Management of missing data: not stated

Selective reporting (reporting bias)High riskComment: no protocol available and duration of recurrences, duration of clinical shedding, and time to first day of shedding not reported

Kinghorn 1985

Methods

Prospective, double-blind, randomized, cross-over trial

1 center in UK

Period of inclusion not stated

Participants

Inclusion criteria

  • Age ≥ 18 years

  • Recurrent genital herpes confirmed by positive viral culture

  • At least 6 recurrences of genital herpes in the preceding year

Exclusion criteria

  • Pregnancy

  • Impaired renal or hepatic function

  • Antiviral or immune stimulatory treatment

Baseline data

Randomly assigned: 40

  • Mean age (extremes), years: 29.4 (18-52)

  • Male/female: 24/16

  • Duration of condition, year (extremes): 3 (1-12)

  • Mean number of recurrences per year (extremes): 12 (6-40)

    • HSV-2 n=37

    • HSV-1 n= 2

    • HSV-1 et HSV-2 n = 1

  • Number of days between last recurrence and initiation of treatment: ≤72hours

Withdrawal: acyclovir 200 x 4/d; placebo

  • Lost of follow-up: 3; 3

  • Failure to complete protocol: 1

  • No recurrence during trial: 1; 1

  • Adverse events: 2; 1

Interventions

Duration of periods of treatment: until first recurrence or maximum 12 weeks

Duration of washout between periods of treatment: not stated

Intervention 1: oral acyclovir 200 mg x 4/d

Intervention 2: placebo

Co-intervention: none

Outcomes

Primary outcome: not stated

Definition of recurrence: not stated

Outcomes

  • Number of participants per group experiencing a recurrence

  • Mean time to first recurrence

  • Adverse events

  • Viral sensitivities to acyclovir

Methods of recurrence assessment

  • Scheduled visit each 4 weeks

  • Unplanned visit motivated by recurrences

  • Specimens for viral culture were taken during visit from urethra for men and from cervix for women and from any clinically suspect genital herpetic lesion

  • Participant diary: tablets taken, symptoms of recurrences

Notes

The study authors thanked the Welcome Foundation for advice and help provided with this trial

Last study author was an employee of Clinical and Applied Research Division , Wellcome Foundation, Beckenham, Kent

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear risk

Quote: "randomised"

Comment: method used to generate the allocation sequence was not reported

Allocation concealment (selection bias)Unclear riskComment: no description of the method used to guarantee allocation concealment
Blinding of participants and personnel (performance bias)
All outcomes
Low risk

Quote: "double blind"

Comment: no precision but as placebo-controlled trial probably done

Blinding of outcome assessment (detection bias)
All outcomes
Low risk

Quote: "double blind"

Comment: no precision but as placebo-controlled trial probably done

Incomplete outcome data (attrition bias)
All outcomes
Unclear risk

Number of randomly assigned participants: 40

Number of analyzed participants for time to first recurrence and rate of recurrence: 28 (16 for ACV first and 12 for PBO first)

Participants withdrawn: 12

  • 6 lost to follow-up (3 ACV, 3 PBO)

  • 1 lack of compliance

  • 2 no recurrences during the first period treatment (1 ACV, 1 PBO)

  • 3 because of adverse events (2 ACV, 1 PBO)

Selective reporting (reporting bias)High riskComment: primary outcome not stated, outcomes not clearly listed in the methods section, no protocol available, no results of viral culture

Kinghorn 1992

Methods

Prospective, double-blind, randomized, parallel-group trial

13 centers in Canada and Europe (Belgium, Germany, United Kingdom)

Period of inclusion not stated

Participants

Inclusion criteria

  • Age ≥ 16 years

  • Recurrent genital herpes confirmed by HSV viral culture during the period of observation of 8 weeks

  • At least 6 recurrences of genital herpes per year

Exclusion criteria

  • Pregnancy

  • Immunosuppression

  • Gastrointestinal malabsorption

  • Gout, hyperuricemia

  • Severe atopic eczema

Randomly assigned: acyclovir (53), placebo (25)

Baseline data

  • Randomly assigned: acyclovir (53), placebo (25)

  • Mean age, years (SD): 33.5 (8.3); 33.7 (10.9)

  • Male/female: 37/16; 15/10

  • Duration of condition, years (extremes): 6.7 (1-25); 6.7 (0-25)

  • Mean number of recurrences past year (SD): 10.9 (5); 11.7 (7.3)

  • Number of days between last recurrence and initiation of treatment: 7 days; 7 days

Withdrawal:not reported

Interventions

Duration of treatment: 24 weeks

Intervention 1: oral acyclovir 400 mg x 2/d + dummy isoprinosine

Intervention 3: dummy acyclovir + dummy isoprinosine

NB : 1 arm excluded (isoprinosine)

Co-intervention: topical acyclovir during recurrence after 2 recurrences

Outcomes

Primary outcome: not stated

Definition of recurrence: recurrence observed by a physician OR reported by particpant self assessment

Outcomes

  • Number of participants with recurrences

  • Number of recurrences

  • Time to first recurrence

  • Mean duration of recurrence

  • Mean duration of symptoms or erythema

  • Adverse events

Methods of recurrence assessment

  • Scheduled visit each month

  • Unplanned visit motivated by recurrences

  • Participant diary: compliance, symptoms, concomitant medication

NotesQuote: "We also acknowledge the support of the department of clinical statistics and data handling, Wellcome Research Laboratories"
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk

Quote: "computer generated randomisation table"

Comment: probably done

Allocation concealment (selection bias)Unclear riskComment: no description of the method used to guarantee allocation concealment
Blinding of participants and personnel (performance bias)
All outcomes
Low risk

Quote: "double blind"

Comment: no precision but as double-dummy controlled trial probably done

Blinding of outcome assessment (detection bias)
All outcomes
Low risk

Quote: "double blind"

Comment: no precision but as double-dummy controlled trial probably done

Incomplete outcome data (attrition bias)
All outcomes
Unclear risk

Randomly assigned: acyclovir (53), placebo (25)

Analyzed for time to first recurrence and recurrence rate: 53/24

Withdrawal: not reported

Management of missing data: not stated

Selective reporting (reporting bias)Unclear riskComment: primary outcome not stated, no protocol available

Kroon 1989

Methods

Prospective, double-blind, randomized, cross-over trial

1 center in Denmark

Period of inclusion not stated

Participants

Inclusion criteria

  • Age ≥ 18 years

  • Recurrent genital herpes confirmed by positive viral culture

  • At least 8 recurrences of genital herpes per year

Exclusion criteria

  • Pregnancy

  • Immunodepression

  • Impaired renal or hepatic function

  • Severe atopic eczema

Baseline data

  • Randomly assigned: 29

  • Mean age, years (extremes):26 (20-43)

  • Male/female: 18/11

  • Duration of condition, years (extremes): 2 (1-6)

  • Median number of recurrences per year (extremes): 13.5 (8-28)

  • Median number of days between last recurrence and initiation of treatment (extremes): 7 (0-48)

Withdrawal:

  • Lost to follow-up: 4

  • No recurrence during trial period: 1

Interventions

Duration of periods of treatment: 12 weeks

Duration of wash-out between periods of treatment: not stated

Intervention 1: oral acyclovir 400 mg x 2 /d

Intervention 2: placebo

Co-intervention: none

Outcomes

Primary outcome: not stated

Definition of recurrence: recurrence observed by a physician OR reported by participant self assessment

Outcomes

  • Number of participants experiencing a recurrence

  • Median time to first recurrence

  • Adverse reaction

  • Viral sensitivities to acyclovir

  • Time to recurrence while receiving ACV first

Methods of recurrence assessment

  • Scheduled visit each 4 weeks

  • Unplanned visit motivated by recurrences

  • Specimens for viral culture were taken during visit from urethra for men and from cervix for women and from any clinically suspect genital herpetic lesion

  • Participant diary: tablets taken, symptoms of recurrences

NotesNo funding reported; however belongs to a group of 5 cross-over trials (Blom 1986; Halsos 1985; Kinghorn 1985; Kroon 1989; Thin 1985) having very similar design of which 4 were declared funded by Wellcome
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear risk

Quote: "randomised"

Comment: method used to generate the allocation sequence was not reported

Allocation concealment (selection bias)Unclear riskComment: no description of the method used to guarantee allocation concealment
Blinding of participants and personnel (performance bias)
All outcomes
Low risk

Quote: "double blind"

Comment: no precision but as placebo-controlled trial probably done

Blinding of outcome assessment (detection bias)
All outcomes
Low risk

Quote: "double blind"

Comment: no precision but as placebo-controlled trial probably done

Incomplete outcome data (attrition bias)
All outcomes
Unclear risk

Number of randomly assigned participants: 29

Number of analyzed participants for time to first recurrence and rate of recurrence: 24 (10 ACV first, 14 PBO first)

Participants withdrawn: 5

  • 4 lost to follow-up

  • 1 no recurrence during trial period

Selective reporting (reporting bias)Unclear riskComment: no stated primary outcome, outcomes not clearly listed in the methods section, no protocol available, unclear results for viral culture, no protocol available

Mattison 1988

Methods

Prospective, double-blind, randomized, parallel-group trial

2 centers in USA

Period of inclusion not stated

Participants

Inclusion criteria

  • Age ≥ 18 years

  • Recurrent genital herpes confirmed by HSV viral culture

  • At least 6 recurrences of genital herpes during the year before enrolment

Exclusion criteria

  • Pregnancy

  • Immunosuppression

  • Antiviral therapy

Baseline data

  • Randomly assigned: acyclovir (52), placebo (52)

  • Mean age, years: 33.9; 33.1

  • Male/female : 32/20; 32/20

  • Number of days between last recurrence and initiation of treatment: 16; 19.6

Withdrawal: acyclovir (5), placebo (23)

Reasons

  • Inefficacy: 1; 6

  • Adverse events: 0; 2

Interventions

Duration of treatment: 1 year

Intervention 1: acyclovir 400 mg x 2/d + dummy intermittent treatment

Intervention 2: dummy acyclovir + dummy intermittent treatment

NB: 1 arm excluded (dummy acyclovir + intermittent treatment)

Co-intervention: none

Outcomes

Primary outcome: not stated

Definition of recurrence: recurrence observed by a physician OR patient self assessment

recurrences during the first week of trial were not taking account for analysis

Outcomes

  • Time to first recurrence

  • Number of recurrences per month

  • Days of active disease

  • Duration of recurrences

  • Median rate of false prodromes

  • Viral culture during recurrences

  • Adverse events

Methods of recurrence assessment

  • Scheduled visit each month

  • Unplanned visit motivated by recurrences within 48 hours after onset of recurrence

  • Participant diary: genital herpes-related symptoms

  • Swab for viral culture during recurrence

Notes1 of 9 study authors were declared as Burroughs Wellcome Company employees
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear risk

Quote: "randomly assigned"

Comment: method used to generate the allocation sequence was not reported

Allocation concealment (selection bias)Unclear riskComment: no description of the method used to guarantee allocation concealment
Blinding of participants and personnel (performance bias)
All outcomes
Low risk

Quote: "double blind"

Comment: no precision but as double-dummy controlled trial probably done

Blinding of outcome assessment (detection bias)
All outcomes
Low risk

Quote: "double blind"

Comment: no precision but as double-dummy controlled trial probably done

Incomplete outcome data (attrition bias)
All outcomes
High risk

Randomly assigned: acyclovir (52), placebo (52)

Analyzed: for time to first recurrence and recurrences rate: 47/29

Withdrawal: acyclovir (5), placebo (23)

Management of missing data: "participants with missing values for particular variables were excluded from analyses that involved those variables"

Comment: High missing data rate in placebo arm. Inbalance in numbers across groups

Selective reporting (reporting bias)High riskComment: primary outcome not stated, outcome mentioned in the methods section fully reported except for viral culture during recurrences, no protocol available

Mertz 1988

Methods

Prospective, double-blind, randomized, parallel-group trial

24 centers in USA and Canada

From March to August 1984

Participants

Inclusion criteria

  • Age ≥ 18 years

  • Recurrent genital herpes for at least 12 month confirmed by HSV viral culture

  • At least 6 recurrences of genital herpes during the year before enrolment

Exclusion criteria

  • Pregnancy/lactation

  • Topical acyclovir for the last episode before enrolment

Baseline data: on analyzed participants acyclovir (575), placebo (571) not on randomly assigned: acyclovir (586), placebo (589)

  • Mean age, years: 33.7; 33.1

  • Male/female: 307/279; 308/263

  • Duration of condition, years: 4.6; 4.7

  • Mean number of recurrences past year (mean): 12.8; 12.7

  • Number of days between last recurrence and initiation of treatment : 23.8; 23.2

Withdrawal: acyclovir (67), placebo (158)

Reasons

  • No follow-up visit: 11; 18

  • Participants moved, or non compliant, or want oral acyclovir: 52; 128

  • Liver abnormalities due to alcohol: 2; 1

  • Pregnancy: 1; 2

  • Ilness or death: 1; 3

  • Adverse events: 0; 6

Times of withdrawal were not reported

Interventions

Duration of treatment: 1 year

Intervention 1: oral acyclovir 400 mg x 2/d

Intervention 2: placebo

Co-intervention: acyclovir 200mg x 5/d during 5 days in case of recurrence

Outcomes

Primary outcome:not stated

Definition of recurrence: recurrence observed by a physician OR reported by participant self assessment

Outcomes

  • Time to first recurrence

  • Frequency of episodes

  • Frequency of false prodromes

  • Changes in laboratory values

  • Duration of first episode after treatment discontinued

Method of recurrences assessment:

  • Scheduled visit each month

  • Unplanned visit motivated by recurrences within 48 hours after onset of recurrence

  • Participant diary: missed capsules, presence of prodromes, lesion pain, oral or genital lesions

Notes

2 of 8 authors were declared as Burroughs Wellcome Company employees

Quote: "This trial was supported by grants from the Burroughs-Wellcome"

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear risk

Quote: "randomised"

Comment: method used to generate the allocation sequence was not reported

Allocation concealment (selection bias)Unclear riskComment: no description of the method used to guarantee allocation concealment
Blinding of participants and personnel (performance bias)
All outcomes
Low risk

Quote: "double blind"

Comment: no precision but as placebo-controlled trial probably done

Blinding of outcome assessment (detection bias)
All outcomes
Low risk

Quote: "double blind"

Comment: no precision but as placebo-controlled trial probably done

Incomplete outcome data (attrition bias)
All outcomes
High risk

Randomly assigned: acyclovir 586, placebo 589

Analyzed

  • For time to first recurrence: 575; 571

  • For rate of recurrence: 519; 431

Withdrawal: acyclovir (67), placebo (158)

Comment: high missing data rate in placebo arm. Inbalance in numbers across groups

Selective reporting (reporting bias)Unclear riskComment: primary outcome not stated, no protocol available

Mertz 1997

Methods

Prospective, double-blind, randomized, parallel-group trial

20 centers in USA

Period of inclusion not stated

Participants

Inclusion criteria

  • Age ≥ 18 years

  • Recurrent genital herpes

  • At least 6 recurrences of genital herpes during the year before enrollment in the absence of antiviral treatment or in any 12-month period in the past 24 months while not receiving antiviral treatment and at least 1 recurrence per month since discontinuing therapy

  • Washout period of 2 months required for oral acyclovir or 14 days for any antiviral treatment

Exclusion criteria

  • Pregnancy/lactation

  • Immunodepression including positive HIV serology

Baseline data: number of randomly assigned: 375

  • Mean age, years: 34

  • Male/female: 0/375

  • Duration of condition, years: not stated

  • Mean number of recurrences per year (median): 7

  • Number of days between last recurrence and initiation of treatment : not reported

Withdrawal: famciclovir 125 mg x 1 (12); famciclovir 125 mg x2 (7); famciclovir 250mg x 1 (14); famciclovir 250 mg x 2(13); famciclovir 500 mg x 1 (7); placebo (10)

Reasons for and times of withdrawal were not reported

Interventions

Duration of treatment: 1 year

Intervention 1: oral famciclovir 125mg x 1/d

Intervention 2: oral famciclovir 125mg x 2/d

Intervention 3: oral famciclovir 250mg x 1/d

Intervention 4: oral famciclovir 250mg x 2/d

Intervention 5: oral famciclovir 500mg x 1/d

Intervention 6: placebo

Co-intervention: none

Outcomes

Primary outcome: time to first recurrence and number of recurrences during treatment (episodes during 120 hours after start of treatment were excluded)

Definition of recurrence: symptomatic episode with lesions (observed by a physician or self reported by the participant) or a positive HSV viral culture or both

Secondary outcomes

  • Adverse events

Methods of recurrence assessment

  • Scheduled visit each month

  • Unplanned visit motivated by recurrences within 24 hours after onset of recurrence

  • Participant diary: compliance, day of onset, and day of healing

NotesQuote: "This trial was funded by SmithKline Beecham Pharmaceuticals"
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear risk

Quote: "subject sequentially allocated a trial number at each centre according to randomisation code"

Comment: method used to generate the allocation sequence was not reported

Allocation concealment (selection bias)Unclear riskComment: no description of the method used to guarantee allocation concealment
Blinding of participants and personnel (performance bias)
All outcomes
Low riskQuote: "tablets containing placebo and those containing doses of famciclovir were identical, no report of compromised blinding by vision or taste"
Blinding of outcome assessment (detection bias)
All outcomes
Low risk

Quote: "double blind"

Comment: probably done

Incomplete outcome data (attrition bias)
All outcomes
Unclear risk

Randomly assigned: 375

Analyzed:

famciclovir125 mg x 1 (60); famciclovir 125 mg x 2 (65); famciclovir 250 mg x 1 (61); famciclovir 250 mg x 2 (64); famciclovir 500 mg x 1 (61); placebo (64)

Withdrawal: famciclovir 125 mg x 1 (12); famciclovir 125 mg x 2 (7); famciclovir 250 mg x 1 (14); famciclovir 250 mg x 2 (13); famciclovir 500mg x 1 (7); placebo (10)

Management of missing data: not stated

Selective reporting (reporting bias)Unclear riskComment: protocol not available; outcomes announced in the methods section were fully reported

Mindel 1984

Methods

Prospective, double-blind, randomized, parallel-group trial

1 center in UK

Period of inclusion not stated

Participants

Inclusion criteria

  • Age ≥ 16 years

  • Recurrent genital herpes with at least 1 confirmed by HSV viral culture during a 3-month observation period

  • At least 4 recurrences of genital herpes per year

  • Washout of 1 month required for antiviral treatment

Exclusion criteria

  • Pregnancy

  • Impaired renal function

  • participants unable to attend at the required interval

Baseline data:

  • Randomly assigned: acyclovir (29), placebo (27)

  • Mean age, years (SE): 31.3 (1.5); 30.4 (1.4)

  • Male/female: 13/16; 9/18

  • Duration of condition, years: not reported

  • Mean number of recurrences past year (mean): not reported

  • Number of days between last recurrence and initiation of treatment: 7days; 7days

Withdrawal: not reported

Interventions

Duration of treatment: 12 weeks

Intervention 1: oral acyclovir 200 mg x 4/d

Intervention 2: placebo

Co-intervention: none

Outcomes

Primary outcome: not stated

Definition of recurrence: unclear

Outcomes

  • % of participants with recurrences

  • Mean number of recurrences per month

  • Time to first recurrence

  • Number of recurrences with positive viral HSV culture

  • Adverse events

  • Compliance

Methods of recurrence assessment

  • Scheduled visit each 2 weeks

  • Unplanned visit motivated by recurrences

  • No participant diary

NotesQuote: "we report the interim results of the trial"
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear risk

Quote: "randomised"

Comment: no precision on the randomization method

Allocation concealment (selection bias)Unclear riskComment: no description of the method used to guarantee allocation concealment
Blinding of participants and personnel (performance bias)
All outcomes
Low risk

Quote: "double blind"

Comment: no precision but as placebo-controlled trial probably done

Blinding of outcome assessment (detection bias)
All outcomes
Low risk

Quote: "double blind"

Comment: no precision but as placebo-controlled trial probably done

Incomplete outcome data (attrition bias)
All outcomes
Unclear risk

Randomly assigned: acyclovir (29), placebo (27)

Analyzed: 29; 27

Withdrawal: not reported

Management of missing data: not stated

Comment:

Selective reporting (reporting bias)Unclear riskComment: no protocol available, no primary outcome stated

Mostow 1988

Methods

Prospective, double-blind, randomized, parallel-group trial

1 center in USA

Period of inclusion not stated

Participants

Inclusion criteria

  • Adults

  • Recurrent genital herpes confirmed by HSV viral culture

  • At least 4 recurrences of genital herpes during the 6 previous months

Exclusion criteria

  • Pregnancy

  • Herpetic lesion at enrolment

Baseline data:

  • Randomly assigned: acyclovir (22); placebo (24)

  • Mean age, years (extremes): 32.7 (25-48); 32.6 (23-48)

  • Male/female : 10/12; 14/10

  • Duration of condition, year (extreme): 18.5 (9-54); 17.1 (7-40)

  • Mean number of recurrences past year (SD): not reported

  • Number of days between last recurrence and initiation of treatment: not reported

Withdrawal: acyclovir (1), placebo (5)

Reasons

  • No follow-up visit: 1; 2

  • Pregnancy: 0; 2

  • Treatment failure: 0; 1

Times of withdrawal were not reported

Interventions

Duration of treatment: 1 year

Intervention 1: oral acyclovir 800 mg x 1/d

Intervention 2: placebo

Co-intervention: acyclovir 200 mg x 5/d during 5 days in case of recurrence

Outcomes

Primary outcome: not stated

Definition of recurrence: clinical recurrence observed by a physician

Outcomes:

Number of participants recurrence free

Number of recurrences per month

Number of recurrences with positive viral culture

Adverse events

Methods of recurrence assessment

Scheduled visit each month

Unplanned visit motivated by recurrences

No participant diary

Notes2 of 4 study authors were declared as Burroughs Wellcome Company employees
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk

Quote: "patients were randomly distributed into treatment group by computer generated code"

Comment: probably done

Allocation concealment (selection bias)Unclear riskComment: no description of the method used to guarantee allocation concealment
Blinding of participants and personnel (performance bias)
All outcomes
Low risk

Quote: "double blind matching placebo"

Comment: no precision but as placebo-controlled trial probably done

Blinding of outcome assessment (detection bias)
All outcomes
Low risk

Quote: "double blind"

Comment: no precision but as placebo-controlled trial probably done

Incomplete outcome data (attrition bias)
All outcomes
High risk

Randomly assigned: acyclovir (22); placebo (24)

Analyzed for rate of recurrence: 21; 19

Withdrawal: acyclovir (1), placebo (5)

Comment: unbalance quantities of missing data across group

Selective reporting (reporting bias)High riskComment: primary outcome not stated, results of viral HSV culture were reported globally for the 2- year trial, including 1 year nonrandomized follow-up, no protocol available

Patel 1997

Methods

Prospective, double-blind, randomized, parallel-group trial

34 centers in USA

From December 1994 to May 1996

Participants

Inclusion criteria

  • Age ≥ 18 years

  • Recurrent genital herpes

  • At least 8 recurrences of genital herpes per year

  • 3-month washout period required for antiviral treatment and at least 1 recurrence during washout period

Exclusion criteria

  • Pregnancy/lactation

  • Immunosuppression

  • impaired renal function

  • Hepatic impairment

  • Hypersensitivity to acyclovir or valacyclovir

  • Participant receiving systemic or topical antiviral or immunomodulatory treatments

Baseline data

  • Randomly assigned: valacyclovir (288); placebo (94)

  • Mean age, years (extremes): 33 (19-79); 34 (19-80)

  • Male/female : 136/152; 39/55

  • Duration of condition, years (extreme): 5.1 (0.1-52.2); 3.8 (0.2-39.8)

  • Mean number of recurrence last year (SD): not reported

  • Number of days between last recurrence and onset of treatment : not reported

Withdrawal: valacyclovir (20), placebo (2)

Reasons

  • Adverse events: 7; 0

  • Other: 13; 2

Times of withdrawal were not reported

Interventions

Duration of treatment: 16 weeks (or until first recurrence)

Intervention 1: oral valacyclovir 500 mg x 1/d

Intervention 2: placebo

Co-interventions: valacyclovir 500 mg x 2/d during 5 days in case of recurrence

Outcomes

Primary outcome: time to first recurrence

Definition of recurrence: clinical recurrence observed by a physician or self reported by participant. Podromal symptoms alone were not sufficient to define a recurrence; a papular vesicular stage is required

Outcomes

  • % of participants recurrence free

  • Adverse events

Methods of recurrence assessment

  • Scheduled visit each 4 weeks

  • Unplanned visit motivated by recurrences within the 24 hours from onset of the recurrence

  • Participant diary: prodromal symptoms, any recurrence of genital symptoms, any concomitant medication, any adverse event

Notes

2 of 4 study authors were declared as Glaxo Wellcome employees

Quotes: "...gratefully acknowledge Glaxo Wellcome for financial support and assistance with the conduct of this trial"

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear risk

Quote: "randomised"

Comment: method used to generate the allocation sequence was not reported

Allocation concealment (selection bias)Unclear riskComment: no description of the method used to guarantee allocation concealment
Blinding of participants and personnel (performance bias)
All outcomes
Low risk

Quote: "double blind"

Comment: no precision but as placebo controlled-trial probably done

Blinding of outcome assessment (detection bias)
All outcomes
Low risk

Quote: "double blind"

Comment: no precision but as placebo controlled-trial probably done

Incomplete outcome data (attrition bias)
All outcomes
High risk

Randomly assigned: valacyclovir (288); placebo (94)

Analyzed for time to first recurrent and rate of recurrence: 288; 94

Withdrawal: valacyclovir (20), placebo (2)

Management of missing data: "For participants who did not reach this endpoint, censored event-free times were calculated to be equal to the last day that no event was recorded"

Comment: unbalanced quantity of missing data across group

Selective reporting (reporting bias)Unclear riskComment: protocol not available, and outcomes announced in the methods section were fully reported

PREV123 2006

Methods

Prospective, open-label, randomized, parallel-group trial

76 centers in France

From September 25, 1997 to September 23, 1998

Participants

Inclusion criteria

  • Adults

  • Recurrent genital herpes confirmed by positive HSV viral culture or by written practitioner's confirmation

  • At least 6 recurrences of genital herpes during the year before enrollment or at least 3 in the past 6 months

  • 6-week washout period required for oral acyclovir and at least 1 recurrence during washout period and 1-week wash-out period for any topical or systemic treatment

Exclusion criteria

  • Pregnancy/lactation

  • Known or suspected immunosuppression

  • Impaired renal function

  • Liver deficiency

  • Recurrence on inclusion visit (included 1 week later after antiviral treatment)

Baseline data

  • Randomly assigned: valacyclovir (122); no intervention (115)

  • Mean age, years (SD): 45.1 (12.4) for all the population

  • Male/female: 58/64; 33/82

  • Duration of condition, years: not reported

  • Mean number of recurrences past year: not reported

  • Number of days between last recurrence and initiation of treatment: not reported

Withdrawal: valacyclovir (16), placebo (8)

Reasons:

  • Adverse events: 6; 1

  • Lack of efficacy: 1; 4

  • Others: 9; 3

Times of withdrawal were not reported

Interventions

Duration of treatment: 6 months

Intervention 1: oral valacyclovir 500 mg x 1/d

Intervention 2: no intervention

Co-intervention: valacyclovir 500 mg x 2/d during 5 days in case of recurrence

Outcomes

Primary outcome: time to first recurrence

Definition of a recurrence: unclear

Secondary outcomes

  • Number of recurrences

  • Frequency and types of adverse events

  • Evolution of quality of life in herpes questionnaire (QLH)

Methods of recurrence assessment: not reported

Notes 
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear risk

Quote: "randomised"

Comment: method used to generate the allocation sequence was not reported

Allocation concealment (selection bias)Unclear riskComment: no description of the method used to guarantee allocation concealment
Blinding of participants and personnel (performance bias)
All outcomes
High risk

Quote: "open-label trial"

Comment: subjective outcome and method of recurrence assessment not reported

Blinding of outcome assessment (detection bias)
All outcomes
High risk

Quote: "open label trial"

Comment: subjective outcome and method of recurrence assessment not reported

Incomplete outcome data (attrition bias)
All outcomes
High risk

Randomly assigned: valacyclovir (122); no intervention (115)

Analyzed for time to first: 122; 115

Withdrawal: valacyclovir (16), placebo (8)

Management of missing data: not stated

Comment: unbalanced quantities of missing data across group

Selective reporting (reporting bias)Low riskComment: protocol not available

Reitano 1998

Methods

Prospective, double-blind, randomized, parallel-group trial

53 centers in USA, Australia, United Kingdom, Denmark

From December 1994 to May 1996

Participants

Inclusion criteria

  • Age ≥ 18 years

  • Recurrent genital herpes confirmed by positive HSV culture or direct antigen test or Tzanck smear or immunofluorescence assay or written confirmation by primary care physician

  • At least 6 recurrences of genital herpes per year

  • 3-month washout period required for antiviral treatment and at least 1 recurrence during washout period

Exclusion criteria

  • Pregnancy/lactation

  • Immunosuppression

  • Impaired renal function

  • Hepatic impairment

  • Hypersensitivity to acyclovir or valacyclovir

  • Participants receiving systemic or topical antiviral or immunomodulatory treatments or probenecid

Baseline data

  • Randomly assigned: valacyclovir 250 x 2 (274); valacyclovir 1000 mg x 1 (269); valacyclovir 500 x 1 (266); valacyclovir 250 x 1 (250); acyclovir 400 x 2 (267); placebo (134)

  • Mean age, years (extremes): 34 (18-85); 34 (31-71); 34(17-74); 34 (18-81); 33(20-73); 33(19-58)

  • Male/female: 131/143; 132/137; 123/143; 142/127; 126/141; 55/79

  • Duration of condition, years: 5.6; 6.1; 4.6; 4.9; 4.6; 5.7

  • Mean number of recurrences past year (SD): not reported

  • Number of days between last recurrence and initiation of treatment: not reported

Withdrawal: for the total included population : 429

Reasons

  • Lost to follow-up: 39

  • Withdrawal of consent: 36

  • Adverse events: 22

  • Protocol violation: 10

  • Other: 322

Times of withdrawal were not reported

Interventions

Duration of treatment: 16 weeks (or until first recurrence)

Intervention 1: oral valacyclovir 250 mg x 2/d

Intervention 2: oral valacyclovir 1000 mg x 1/d

Intervention 3: oral valacyclovir 500 mg x 1/d

Intervention 4: oral valacyclovir 250 mg x 1/d

Intervention 5: oral acyclovir 400 mg x 2/d

Intervention 6: placebo

Co-intervention: valacyclovir 500 mg x2/d during 5 days in case of recurrence

Outcomes

Primary outcome: time to first recurrence

Definition of recurrence: clinical recurrence observed by a physician or self reported by participant. Lesions at papular vesicular stage is required.

Outcomes

  • Time to second recurrence

  • % of participants free of recurrence at 12 month

  • Adverse events

Methods of recurrence assessment

  • Scheduled visit each month

  • Unplanned visit motivated by recurrences within the 24 hours from onset of the recurrence

  • Participant diary: any recurrence, use of concomitant medication, any adverse event

Notes

Some study authors were declared as Burroughs Wellcome Company employees.

Quote: "financial support: Glaxo Wellcome Research and Development"

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear risk

Quote: "randomised"

Comment: method used to generate the allocation sequence was not reported

Allocation concealment (selection bias)Unclear riskComment: no description of the method used to guarantee allocation concealment
Blinding of participants and personnel (performance bias)
All outcomes
Low risk

Quote: "double blind"

Comment: no precision but as placebo-controlled trial probably done

Blinding of outcome assessment (detection bias)
All outcomes
Low risk

Quote: "double blind"

Comment: no precision but as placebo controlled trial probably done

Incomplete outcome data (attrition bias)
All outcomes
Unclear risk

Randomly assigned: valacyclovir 250 x 2 (274); valacyclovir 1000 x 1 (269); valacyclovir 500 x 1 (266); valacyclovir 250 x 1 (250); acyclovir 400 x 2 (267); placebo (134)

Analyzed for time to first recurrence and rate of recurrences :valacyclovir 250x2 (274); valacyclovir 1000x1 (269);valacyclovir 500x1 (266);valacyclovir 250x1 (250); acyclovir 400x2 (267); placebo (134)

Withdrawal: for the total included population: 429

Management of missing data: "for participants who did not reach this end point censored event-free times were calculated to be equal to the number of days until the last day on which no event was recorded" An analysis of crude proportions of participants recurrence free at 12 months was conducted to complement the survival analysis. Participants who withdrew from the trial before 12 months with an unknown recurrence status were counted in the analysis as having had a recurrence during the trial

Comment: high rate of missing data

Selective reporting (reporting bias)Unclear riskComment: protocol not available and time to first recurrence fully reported

Romanowski 2003

Methods

Prospective, open label, randomized, cross-over trial

16 centers in Canada

Period of inclusion not stated

Participants

Inclusion criteria

  • Age ≥ 18 years

  • Documented history (method of documentation required not stated) of genital herpes

  • At least 4 to 9 recurrences past year or 2 to 4 past 6 months

  • At least 1 recurrence required within 16 weeks between withdrawal of suppressive therapy and time of inclusion

  • Washout duration of previous antiviral treatment: 1 month

Exclusion criteria

  • Pregnancy

  • Lactation

  • Immunosuppression

  • More than 10 recurrences per year

  • Ocular HSV

  • Creatinine clearance < 15 mL/min

  • ASAT or ALAT > 3 N

  • Intolerance to acyclovir 

  • Gastrointestinal dysfunction

  • Medication interfering with drug absorption or disposition

Baseline data: data on analyzed participants n = 202; not on randomly assigned n = 225

  • Mean age, years (extremes): 32

  • Male/female: 78/124

  • Duration of condition, years: 4.7

  • Mean number of recurrences past year (extremes): 5.9 (SD 2.1)

  • Median number of days between last recurrence and initiation of treatment: not stated

Withdrawal

  • Adverse events: 6

  • Consent withdrawal: 16

  • Lost to follow-up: 7

  • Protocol violation: 2

  • Other reason: 9

Interventions

Duration of periods of treatment: 24 weeks

Duration of washout between periods of treatment: not stated

Intervention 1: oral valacyclovir 500 mg x 1/day

Intervention 2: no treatment

Co-intervention: valacyclovir 500 x 2/d 5days

Outcomes

Primary outcome: participant treatment preference (GHERPTSQ)

Definition of recurrence:clinical recurrence participant reported

Secondary outcomes

  • Quality of life (RGHQoL)

  • Adverse events

  • Herpes recurrence date on diary

  • Particpant satisfaction with treatment

Methods of recurrence assessment:

Scheduled visit each 4 weeks

Unplanned visit motivated by recurrences

No specimen for viral culture or PCR

Participant diary: trial medication taken, date and time of onset of recurrence, date on which recurrence ended, adverse events, change in nontrial medication

Notes2 of 3 authors were employees of GlaxoSmithKline Company
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskQuote: "treatment order was randomly allocated by the PROC PLAN program within SAS software"
Allocation concealment (selection bias)Unclear riskComment: no description of the method used to guarantee allocation concealment
Blinding of participants and personnel (performance bias)
All outcomes
High risk

Quote: "open label"

Comment: subjective outcome particpant reported

Blinding of outcome assessment (detection bias)
All outcomes
High risk

Quote: "open label"

Comment: subjective outcome participant reported

Incomplete outcome data (attrition bias)
All outcomes
High risk

Number of randomly assigned participants: 225

Number of analyzed participants for time to first recurrence and rate of recurrence: 202 (102 VCV first, 100 PBO first)

Participants withdrawn: 40

  • AE 6

  • Consent withdrawn 16

  • Lost to follow-up 7

  • Protocol violation 2

  • Other reasons 9

Among the 40 withdrawn participants, 23 dropped out in the first period and were excluded from the analysis

Comment: high rate of consent withdrawal possibly linked to the intervention in this unblinded trial

Selective reporting (reporting bias)High riskComment: no protocol available. No results per period for rate of recurrence confirmed clinically. Time to first recurrence: results only on first period

Sacks 1988

Methods

Prospective, double-blind, randomized, parallel-group trial

1 center in Canada

Period of inclusion not stated

Participants

Inclusion criteria

  • Age ≥ 18 years

  • Recurrent genital herpes confirmed by positive HSV culture

  • At least 6 recurrences of genital herpes in the past 6 months and initial episode at least 12 months before enrollment

  • 3-month washout period required for antiviral treatment and at least 1 recurrence during washout period

Exclusion criteria

  • Pregnancy

  • Impaired renal function

Baseline data: (reported for the analyzed population (24; 23); not on randomly assigned: acyclovir (25); placebo (25)

Mean age, years: not reported

Male/female: for the total population: 27/20

Duration of condition, years (median): not reported

Mean number of recurrence last 6 months (SD): 11.1(5.3); 9.3(3.6)

Number of days between last recurrence and initiation of treatment: not reported

Withdrawal: for the total included population : acyclovir (1); placebo (2)

Reasons

  • Personal reason : 1; 0 (dropout at day 6)

  • Recurrence at inclusion: 0; 2

Times of withdrawal were not reported

Interventions

Duration of treatment: 6 months

Intervention 1: oral acyclovir 200 mg x 3/d

Intervention 2: placebo

Co-intervention: acyclovir 200 mg x 5/d during 5 days in case of recurrence

Outcomes

Primary outcome: not stated

Definition of recurrence: clinical recurrence observed by a physician or self reported by a participant

Outcomes

  • Time to second recurrence

  • % of participants free of recurrence

  • Frequency, rate, duration of episodes

  • Number of lesions per episode

  • Nonlesional episodes during treatment

  • Number of virally confirmed lesions (number of positive cultures/number of observed recurrences)

  • Adverse events

Methods of recurrence assessment

  • Scheduled visit each month

  • Unplanned visit motivated by recurrences

  • Participant diary: stage of each recurrence from prodrome to healing

Notes

Some study authors were declared as Burroughs Wellcome Company employees

Quote: "This trial was supported in part by a grant-in-aid from Burroughs Wellcome Company"

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear risk

Quote: "randomised"

Comment: method used to generate the allocation sequence was not reported

Allocation concealment (selection bias)Unclear riskComment: no description of the method used to guarantee allocation concealment
Blinding of participants and personnel (performance bias)
All outcomes
Low risk

Quote: "double blind"

Comment: no precision but as placebo-controlled trial probably done

Blinding of outcome assessment (detection bias)
All outcomes
Low risk

Quote: "double blind"

Comment: no precision but as placebo-controlled trial probably done

Incomplete outcome data (attrition bias)
All outcomes
Low risk

Randomly assigned: acyclovir (25); PBO (25)

Analyzed for time to first recurrence and rate of recurrence : 24; 23

Withdrawal: acyclovir (1); placebo (2)

Selective reporting (reporting bias)Unclear riskComment: no protocol available, primary outcome not stated

Sacks 2004

Methods

Prospective, double-blind, randomized, parallel-group trial

1 center in Canada

Period of inclusion not stated

Participants

Inclusion criteria

  • Women

  • Age ≥ 18 years

  • Recurrent HSV 2 genital herpes confirmed by HSV2 positive serology or positive HSV culture

  • At least 3 recurrences of genital herpes in the past 6 months

  • 2-weeks washout period required for antiviral treatment and at least 1 recurrence/mo during washout period

Exclusion criteria

  • Pregnancy/lactation

  • Immunosuppression

  • Impaired renal function

  • History or clinical suspicion of hepatic, cardiac, gastrointestinal, or hematological dysfunction

  • Genital tract disorder, active genital herpes at time of enrollment

  • Participants receiving immunomodifying treatment, corticosteroids, or other investigational drug

  • History of recent chronic alcoholism or drug abuse

  • Hypersensitivity to acyclovir

Baseline data:

-Randomly assigned: famciclovir 125 mg x 3 (60); famciclovir 250 mg x 2 (60); placebo (58)

-Mean age, years (extremes): 32.4 (20-61); 32.1 (22-50)

-Male/female: 0/60; 0/60; 0/58

-Duration of condition: not reported

-Mean number of recurrences past 6 months: not reported

-Number of days between last recurrence and initiation of treatment: not reported

Withdrawal: not reported

Times of and reasons for withdrawal were not reported

Interventions

Duration of treatment: 4 months

Intervention 1: oral famciclovir 125 mg x 3/d

Intervention 2: oral famciclovir 250 mg x 2/d

Intervention 3: placebo

Co-intervention: none

Outcomes

Primary outcome: % of days of asymptomatic viral shedding from any genital site

Definition of recurrence: clinical recurrence observed by a physician or self reported by a participant

Secondary outcomes

  • % of days of asymptomatic viral shedding from vulvar and from vaginal sites

  • Time to first asymptomatic HSV-positive culture

  • Time to first HSV-positive culture

  • % of days of symptomatic HSV-positive shedding from any genital site

  • Adverse events

Methods of recurrence assessment

  • Scheduled visit each 28 days

  • Unplanned visit motivated by recurrences

  • Participant diary: any symptoms of genital herpes

  • Shedding for viral culture: daily swab (1 vaginal; 1 external) and 1 extra sampling at recurrence site

NotesQuote: "financial support: SmithKline Beecham"
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear risk

Quote: "randomised"

Comment: method used to generate the allocation sequence was not reported

Allocation concealment (selection bias)Unclear riskComment: no description of the method used to guarantee allocation concealment
Blinding of participants and personnel (performance bias)
All outcomes
Low risk

Quote: "double blind "

Comment: no precision but as placebo- and double-dummy -ontrolled trial probably done

Blinding of outcome assessment (detection bias)
All outcomes
Low risk

Quote: "double blind"

Comment:no precision but as placebo- and double-dummy-controlled-trial probably done

Incomplete outcome data (attrition bias)
All outcomes
Unclear risk

Randomly assigned: famciclovir 125 x 3 (60); famciclovir 250 x 2 (60); placebo (58)

Analyzed

  • For total number of patient with at least one virologically: not stated

  • For shedding: 60, 59, 58

Withdrawal: not reported

Management of missing data: not stated

Selective reporting (reporting bias)High riskComment: no protocol available, no primary outcome stated, number of clinically observed recurrences not reported

Sekhin 2004

Methods

Prospective, double-blind, randomized, parallel-group trial

Center not sated

Period of inclusion: 1999-

Participants

Inclusion criteria

  • Both sexes

  • Age: not stated

  • HSV-2+/HSV-1+ or - serostatus determined with Western blot technique

  • At least 4 recurrences per year

Exclusion criteria

  • Not stated

Baseline data:

-Randomly assigned: valacyclovir 500 (22); placebo (20)

-Mean age, years (extremes): NS (18-55)

-Male/female: NS

-Duration of condition: 1-20 years

-Mean number of recurrences per year: 4-9

-Number of days between last recurrence and initiation of treatment: not reported

Withdrawal: not reported

Times of and reasons for withdrawal were not reported

Interventions

Duration of treatment: 4 months

Intervention 1: oral valacyclovir 500 mg/d

Intervention 2: placebo

Co-intervention: valacyclovir 500 mg 2x/d for 5 days

Outcomes

Primary outcome: not stated

Definition of recurrence: not stated

Outcomes

  • Number of participants without recurrence

  • Number of recurrences

Methods of recurrence assessment

  • Not stated

Notes 
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear risk

Quote: "randomized"

Comment: method used to generate the allocation sequence was not reported

Allocation concealment (selection bias)Unclear riskComment: no description of the method used to guarantee allocation concealment
Blinding of participants and personnel (performance bias)
All outcomes
Low risk

Quote: "double blind"

Comment: no precision but as placebo-controlled trial probably done

Blinding of outcome assessment (detection bias)
All outcomes
Low risk

Quote: "double blind"

Comment: no precision but as placebo-controlled trial probably done

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskQuote: no information on withdrawal
Selective reporting (reporting bias)Unclear riskComment: no protocol available, no primary outcome stated

Straus 1984

Methods

Prospective, double-blind, randomized, parallel-group trial

Number of center not reported, USA

Period of inclusion not stated

Participants

Inclusion criteria

  • Women

  • Age ≥ 18 years and ≧50 years

  • Recurrent genital herpes confirmed by positive HSV culture

  • At least 12 recurrences of genital herpes in the previous year and for at least 1 year

Exclusion criteria

  • Pregnancy

Baseline data

  • Randomly assigned: acyclovir (17); placebo (18)

  • Mean age, years (SD): 31.7 (1.2); 32.3 (1.3)

  • Male/female: 8/9; 9/9

  • Duration of condition, years, (DS): 4.1 (0.7); 5.2 (0.9)

  • Mean number of recurrences per year (SD): 14.6 (1.8); 15.3 (1.8)

  • Mean number of recurrences last year (SD): 17.6 (2.4); 16.3 (2)

  • Number of days between last recurrence and onset of treatment (SD): 18.9 (5.2); 21.3 (5.3)

Withdrawal: acyclovir (1), placebo (2)

  • Times and reasons

  • Breast cyst at day 28: 1

  • Error of treatment allocation at month: 1

  • Stop treatment before 4 months: 1

Interventions

Duration of treatment: 125 days or until first recurrence

Intervention 1: oral acyclovir 200 mg x 3/d

Intervention 2: placebo

Co-intervention: none

Outcomes

Primary outcome: not stated

Definition of recurrence: clinical recurrence observed by a physician or clinical impression less compelling with positive viral culture

Outcomes

  • Number of clinical or virological recurrences

  • Number of virologically confirmed recurrences

  • Duration of treatment

  • Compliance

  • Adverse events

Methods of recurrence assessment

  • Scheduled visit every 2 weeks

  • Unplanned visit motivated by recurrences within 24 hours from onset of the recurrence

  • No patient diary

  • Shedding for viral culture: daily swab (1 vaginal; 1 external) and 1 extra sampling at recurrence site

NotesSome study authors were declared as Burroughs Wellcome Company employees
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear risk

Quote: "balance randomisation scheduled with stratification by sex"

Comment: method used to generate the allocation sequence was not reported

Allocation concealment (selection bias)Unclear riskComment: no description of the method used to guarantee allocation concealment
Blinding of participants and personnel (performance bias)
All outcomes
Low risk

Quote : "double blind, coded medication"

Comment: no precision but as placebo-controlled trial probably done

Blinding of outcome assessment (detection bias)
All outcomes
Low risk

Quote : "double blind, coded medication"

Comment: no precision but as placebo-controlled trial probably done

Incomplete outcome data (attrition bias)
All outcomes
Low risk

Randomly assigned: acyclovir (17); placebo (18)

Analyzed for time to first recurrence and rate of recurrence: 16; 16

Withdrawal: acyclovir (1), placebo (2)

Selective reporting (reporting bias)Unclear riskComment: no primary outcome stated, no protocol available

Thin 1985

Methods

Prospective, double-blind, randomized, cross-over trial

5 centers in United kingdom

Period of inclusion not stated

Participants

Inclusion criteria

  • Age ≥ 16 years

  • Recurrent genital herpes confirmed by positive viral culture

  • At least 8 recurrences of genital herpes per year

Exclusion criteria

  • Pregnancy

Baseline data: on analyzed participants n = 88; not on randomly assigned n = 111

  • Mean age, years (extremes): 31 (20-61)

  • Male/female: 64/24

  • Duration of condition, year: not stated

  • Median number of recurrences past year (extremes): 12.6 (8-20)

  • Median number of days between last recurrence and initiation of treatment: 0

Withdrawal: acyclovir 200 mg x 4/d; placebo

  • 10 failed to complete first period

    • Failed to return first visit: 2; 2

    • Adverse events: 0; 2

    • Dropped out: 1, 1

    • Error: 0, 1

    • Cross-over prematurely

  • 10 failed to complete second period: 6; 4

  • 3 inadequate compliance: 1; 2

Interventions

Duration of periods of treatment: until first recurrence or maximum 12 weeks

Duration of washout between periods of treatment: not stated

Intervention 1: oral acyclovir 200 mg x 4 /d

Intervention 2: placebo

Co-intervention: none

Outcomes

Primary outcome: not stated

Definition of recurrence: recurrence defined as symptoms such as local itch and irritation and root pain plus the development of typical lesions with vesicles, erosions and crusting as well as erythema observed by a physician OR by particpant self assessment

Outcomes

  • Number of participants with a recurrence

  • Mean time to appearance of first recurrence

  • Symptoms and signs in the absence of recurrences

  • Adverse events

  • Time to first recurrence after cessation of treatment

Method of recurrences assessment:

  • Scheduled visit each 4 weeks

  • Unplanned visit motivated by recurrences

  • Specimens for viral culture were taken at initial visit and from all subsequent lesion

  • Participant diary: medication, symptoms of recurrence

NotesTwo study authors were employees of Wellcome Research Laboratories, Beckenham, Kent, England
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear risk

Quote: "randomised stratified on gender and centre"

Comment : method used to generate the allocation sequence was not reported

Allocation concealment (selection bias)Unclear riskComment: no description of the method used to guarantee allocation concealment
Blinding of participants and personnel (performance bias)
All outcomes
Low risk

Quote: "double blind"

Comment: no precision but as placebo-controlled trial probably done

Blinding of outcome assessment (detection bias)
All outcomes
Low risk

Quote: "double blind"

Comment: no precision but as placebo)controlled trial probably done

Incomplete outcome data (attrition bias)
All outcomes
High risk

Number of randomly assigned participants: 111

Number of analyzed participants for time to first recurrence and rate of recurrence: 88 (47 PBO first, 41 ACV first)

Participants withdrawn: 23

10 failed to complete first period:

  • Failed to return first visit: 2 ACV; 2 PBO

  • Adverse events: 2 PBO

  • Dropped out: 1 ACV, 1 PBO

  • Failed to complete more than 8 weeks because of an error

  • Cross-over prematurely: 1 ACV

10 failed to complete second period: 6; 4

3 inadequate compliance: 1 ACV; 2 PBO

Selective reporting (reporting bias)High riskComment: no protocol available, no primary outcome stated, outcomes not clearly listed in the methods section, no protocol available. Rate of recurrence confirmed clinically: fully reported. Rate of recurrence confirmed virologically: unclear results. Time to first recurrence: reported without P or CI

Velasco 1991

Methods

Prospective, double-blind, randomized, parallel-group trial

1 center in Chile

From June 1985 to December 1988

Participants

Inclusion criteria

  • Recurrent genital herpes confirmed by positive HSV culture

  • More than 4 recurrences of genital herpes in the previous year

  • Heterosexuallity

Exclusion criteria

  • Pregnancy

  • Concomitant disease or treatment, modifying immune response

Baseline data

  • Randomly assigned: acyclovir (29); placebo (26)

  • Mean age, years (extreme): 38 (21-57)

  • Male/female : 36/19

  • Duration of condition: not reported

  • Mean number of recurrences per year (extreme): 12 (4-12); 9 (4-18)

  • Number of days between last recurrence and initiation of treatment: not reported

Withdrawal: acyclovir(3), placebo (1)

Reasons

  • Moved from the area: 2; 0

  • Psychosis: 1; 0

  • Treatment failure: 0; 1

Times of withdrawal were not reported

Interventions

Duration of treatment: 6 months

Intervention 1: oral acyclovir 400 mg x 2/d

Intervention 2: placebo

Both groups were treated with acyclovir 200 mg x 5/d ,10 days just after enrollment

No intervention: none

Outcomes

Primary outcome: not stated

Definition of recurrence: clinical recurrence (physician observed or participant self reported: not stated) and recurrence with positive viral culture were reported separately

Outcomes

  • Number of patient with clinical recurrences

  • Number of participants with virologically confirmed recurrences

Methods of recurrence assessment

  • Scheduled visit : day 3,5,7,10,15,30,60,90,180

  • No unplanned visit motivated by recurrences

  • No participant diary

NotesQuote: "partially financed by Wellcome Foundation"
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear risk

Quote: "patient included in each group ap hazard"

Comment : method used to generate the allocation sequence was not reported

Allocation concealment (selection bias)Unclear riskComment: no description of the method used to guarantee allocation concealment
Blinding of participants and personnel (performance bias)
All outcomes
Low risk

Quote: "double blind, same appearance placebo"

Comment: no precision but as placebo-controlled trial probably done

Blinding of outcome assessment (detection bias)
All outcomes
Low risk

Quote: "double blind, same appearance placebo"

Comment: no precision but as placebo-controlled trial probably done

Incomplete outcome data (attrition bias)
All outcomes
Low risk

Randomly assigned: acyclovir (29); placebo (26)

Analyzed for rate of recurrence : 26, 25

Withdrawal: acyclovir (3), placebo (1)

Selective reporting (reporting bias)Unclear riskComment: no outcome available, no primary outcome stated, no protocol available, and total numbers of clinical and virologically confirmed recurrences not reported

Wald 2006 study 1

Methods

Prospective, double-blind, randomized, parallel-group trial

22 centers in USA and Canada

From February to October 1997

Participants

Inclusion criteria

  • Age ≥ 18 years

  • Recurrent genital herpes confirmed by positive HSV2 culture or associated with positive HSV-2 or HSV-1 serology

  • At least 6 recurrences of genital herpes in the previous year or history of frequent recurrences before initiation of suppressive antiviral therapy

  • 2-month washout period required for antiviral treatment

Exclusion criteria

  • Pregnancy

  • Human immunodeficiency virus (HIV) infection

  • Impaired renal function

  • Liver disease

  • Other investigational drugs

  • Sensitivity to nucleoside analogues

  • History of acyclovir- or penciclovir- resistant HSV

Baseline data:

  • Randomly assigned: famciclovir (159); valacyclovir (161)

  • Mean age, years (extreme): 36; 38

  • Male/female: 64/95; 71/90

  • Duration of condition, yeras (mean): 13.17; 13.42

  • Mean number of recurrences past year: 9; 9.1

  • Number of days between last recurrence and initiation of treatment: not reported

Withdrawal:famciclovir (13); valacyclovir (19)

Times of and reasons for withdrawal were not reported

Interventions

Duration of treatment: 16 weeks

Intervention 1: oral famciclovir 250 mg x 2/d + dummy valacyclovir

Intervention 2: oral valacyclovir 500 mg x 1/d + dummy famciclovir

Co-intervention: none

Outcomes

Primary outcome: % of particpant who had a clinically confirmed recurrence

Definition of recurrence: clinical recurrence confirmed by a clinician

Secondary outcomes

  • % of participants who had a virologically confirmed recurrence

  • Time to first clinically confirmed recurrence

  • Time to first virologically confirmed recurrence

Recurrences within first 5 days were excluded from efficacy analysis

Methods of recurrence assessment

  • Scheduled visit every 4 weeks

  • Unplanned visit motivated by recurrences within 24 hours from onset of the recurrence

  • No participant diary

  • Swab of lesion for HSV culture

NotesQuote: "funding for the trial was provided by Smithkline Beecham"
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear risk

Quote: "randomised"

Comment: method used to generate the allocation sequence was not reported

Allocation concealment (selection bias)Unclear riskComment: no description of the method used to guarantee allocation concealment
Blinding of participants and personnel (performance bias)
All outcomes
Low risk

Quote: "double blind"

Comment: no precision but as double dummy-controlled trial probably done

Blinding of outcome assessment (detection bias)
All outcomes
Low risk

Quote: "double blind; double dummy; all laboratory studies were performed by technicians masked to the patient code or therapy"

Comment: probably done

Incomplete outcome data (attrition bias)
All outcomes
Unclear risk

Randomly assigned: famciclovir (159); valacyclovir (161)

Analyzed for time to first recurrence: 157; 158

Withdrawal: famciclovir (13); valacyclovir (19)

Management of missing data: not stated

Selective reporting (reporting bias)Unclear riskComment: no protocol available

Wald 2006 study 2

Methods

Prospective, double-blind, randomized, parallel-group trial

3 centers in USA and Canada

From December to May 1998

Participants

Inclusion criteria

  • Recurrent genital herpes

  • At least 6 recurrences of genital herpes in the previous year or in the 12 months before initiation of suppressive antiviral therapy

  • 14-day washout period required for antiviral treatment

Exclusion criteria

  • Liver or kidney disease

  • Human immunodeficiency virus (HIV) infection

  • Participants receiving immunosuppressive therapy

  • Other investigational new drugs

Baseline data: Randomly assigned: famciclovir (34); valacyclovir (36)

  • Mean age, years (extreme): 36; 34.9

  • Male/female : 11/23; 9/27

  • Duration of condition, year (mean): 6.9;7.3

  • Mean number of recurrences past year: 9.5; 9.8

  • Number of days between last recurrence and initiation of treatment: not reported

Withdrawal: famciclovir (1); valacyclovir (3)

Reasons

  • Withdrawal of consent or lost to follow-up: 3

  • Adverse events: 1

Times of withdrawal were not reported

Interventions

Duration of treatment: 10 weeks

Intervention 1: oral famciclovir 250 mg x 2/d + dummy valacyclovir

Intervention 2: oral valacyclovir 500 mg x 1/d + dummy famciclovir

Co-interventions: none

Outcomes

Primary outcome: % of days with positive HSV PCR

Secondary outcomes

  • % of subclinical shedding days

  • % of lesional HSV PCR positive days

  • Time to first shedding

  • Time to first subclinical shedding

  • Time to first lesional shedding

Methods of recurrence assessment

  • Scheduled visit at 2, 4, 7, 10 weeks

  • No unplanned visit motivated by recurrences

  • No participant diary

  • Shedding (PCR): daily swab from genital secretions

NotesQuote: "funding for the trial was provided by SmithKline Beecham"
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear risk

Quote: "randomised"

Comment : method used to generate the allocation sequence was not reported

Allocation concealment (selection bias)Unclear riskComment: no description of the method used to guarantee allocation concealment
Blinding of participants and personnel (performance bias)
All outcomes
Low risk

Quote: "double blind"

Comment: no precision but as double dummy-controlled trial probably done

Blinding of outcome assessment (detection bias)
All outcomes
Low risk

Quote: "double blind...all laboratory studies were performed by technicians masked to the patient code or therapy"

Comment: probably done

Incomplete outcome data (attrition bias)
All outcomes
Low risk

Randomly assigned: famciclovir (34); valacyclovir (36)

Analyzed for shedding: 34; 36

Withdrawal: famciclovir (1); valacyclovir (3)

Management of missing data: not stated

Selective reporting (reporting bias)Unclear riskComment: no protocol available

Characteristics of excluded studies [ordered by study ID]

StudyReason for exclusion
Bartlett 2008Comparator: intermittent treatment
Celum 2008No minimum number of recurrences required for inclusion
Corey 2004bNo minimum number of recurrences per year required for inclusion
Fife 2007Comparator: intermittent treatment
Fuchs 2010No minimum number of recurrences required for inclusion
Gold 1988Ancillary trial (results on effects of treatment on HSV-2 antibodies) from an RCT suppressive ACV vs PBO; no reference to a published version of the results of the main trial
Gupta 2004No minimum number of recurrences required for inclusion
HS240018Trial stopped after 6 inclusions (prematurely terminated following discussion with FDA concerning trial objectives and endpoints)
HS240021Trial stopped after 2 inclusions (prematurely terminated following discussion with FDA concerning trial objectives and endpoints)
Johnston 2011Multi-dose VCV trial without placebo group
Johnston 2012No minimum number of recurrences required for inclusion
Leone 2007no minimum number of recurrences required for inclusion
Mindel 1988Multidose ACV trial without placebo group
Mindel 1989Comparison ACV against isoprinosine
Schiffer 2011Pooled data of 3 studies : 2 excluded (Gupta 2004, Wald 1997) for no minimum number of recurrences required for inclusion and one included (Wald 2006)
Strachan 2011Ancillary trial of Wald 1996 excluded for no minimum number of recurrences required for inclusion
Straus 1986Multi-dose acyclovir trial without placebo group
Wald 1996No minimum number of recurrences required for inclusion
Watson-Jones 2008No minimum number of recurrences required for inclusion

Characteristics of studies awaiting assessment [ordered by study ID]

Douglas 1988

MethodsRandomized, double blind, parallel-group trial
Participants33 participants with at least 4 recurrences per year
InterventionsAcyclovir 200 mg x 5/d vs acyclovir 400 mg x 2/d vs placebo, during 6 months
OutcomesSperm count, mobility, and morphology; mean number of recurrences per month in each group
Notes

Outcome, number of participants with at least 1 recurrence, is not available

Email to the study authors

Frenkel 1989

MethodsRandomized , double blind, parallel-group trial
Participants50 participants with at least 6 recurrences per year
InterventionsAcyclovir unclear if 200 mg or 400 mg twice a day vs placebo during 1 year
OutcomesHSV-2 specific lymphocyte proliferation, mean number of recurrences per year in each group
Notes

Outcome, number of participants with at least 1 recurrence, is not available,

Email to the study authors

Mastrolorenzo 1991

MethodsRandomized, parallel-group trial
Participants68 participants with 6 or more recurrences during the 6 previous months
InterventionsAcyclovir 200 mg x 5/d in case of recurrences during ? vs acyclovir 400 mg x 2/d 12 months
Outcomes% of participants with fewer than 5 recurrences per month in each group
Notes

Outcome, number of particpants with at least 1 recurrence, is not available

Need for more information for evaluation of risk of bias

Email to the study authors

Tyring 2003

Methods2 prospective, double-blind, randomized, parallel-group trial
Participants468 participants with 6 or more recurrences per year
InterventionsFCV 250 mg x 2/d vs placebo 52 weeks
OutcomesProportion of participants free from recurrence
NotesPooled results from 2 RCTs (one included: Diaz-Mitoma 1998), but the results of the other trial not found. We are waiting for an answer from the study authors

Wu 2000

MethodsProspective, double-blind, randomized, parallel-group trial
Participants101 participants
InterventionsVCV 300 mg BID 1 week vs VCV 300 mg BID one month vs ACV 200 mg 5qd 1 month
OutcomesUnclear
NotesFull paper published in a Chinese journal was not found

Zhang 2011

MethodsProspective, double-blind, randomized, parallel-group trial
ParticipantsRecurrent genital herpes
InterventionsACV 400 mg x 2/d 6 months, VCV 500 mg x 2/d 3 months, 500 x 1/d 2 months, 500 mg/eod 1 month
OutcomesNumber of recurrences before and after treatment
NotesFull paper published in a Chinese journal was not found

Characteristics of ongoing studies [ordered by study ID]

NCT01658826

Trial name or title Safety and efficacy comparator trial of a new drug against genital herpes
MethodsPhase 2
Participants
  • Adult, immunocompetent men and women

  • Seropositive for herpes simplex virus type 2 (HSV-2)

  • History of recurrent episodes (>=4 to <=9) of genital herpes for at least 12 months

Expected enrollment: 90

InterventionsAIC316 100 mg once daily vs valacyclovir 500 mg once daily for 28 days
OutcomesWithin-participant genital HSV mucocutaneous shedding rate: number of HSV positive swabs per particpant relative to the total number of swabs collected per particpant
Starting dateOctober 2012
Contact information 
Notes 

Ancillary