Fatal outcome after reactivation of inherited chromosomally integrated HHV-6A (iciHHV-6A) transmitted through liver transplantation.

HHV-6A and HHV-6B are found as inherited and chromosomally integrated forms (iciHHV-6A and -6B) into all germinal and somatic cells and vertically transmitted in a Mendelian manner in about 1% of the population. They were occasionally shown to be horizontally transmitted through hematopoietic stem cell transplantation. Here, we present a clinical case of horizontal transmission of iciHHV-6A from donor to recipient through liver transplantation. Molecular analysis performed on three viral genes (7.2 kb) in the recipient and donor samples supports transmission of iciHHV-6A from the graft. Transmission was followed by reactivation, with high viral loads in several compartments. The infection was uncontrollable, leading to severe disease and death, despite antiviral treatments and the absence of resistance mutations. This case highlights the fact that physicians should be aware of the possible horizontal transmission of iciHHV-6 and its consequences in case of reactivation in immunocompromised patients.

Presence of HHV-6 genome in spermatozoa in a context of couples with low fertility: what type of infection?

Human herpesvirus-6 (HHV-6) is a betaherpesvirus whose genome may integrate into human chromosomes. Chromosomally integrated HHV-6 (ciHHV-6) may be transmitted vertically from parents to children. HHV-6 DNA has been detected in semen, but its integrated or extrachromosomal status has not yet been characterised. The aim of this study was to determine the prevalence of HHV-6 DNA and to search for ciHHV-6 forms in spermatozoa purified from semen obtained from subjects explored for low fertility. A total of 184 sperm samples were purified using PureSperm(®) . HHV-6 viral load and species identification were performed by real-time polymerase chain reaction. Of 179 sperm specimens analysed, three were positive for HHV-6 (1.7%). Two samples (1.1%) had viral loads of 680 232 and 2 834 075 copies per million spermatozoa, compatible with loads expected for a ciHHV-6 form. The viral load of the third positive sample (73 684 copies per million spermatozoa) was lower than would be expected for ciHHV-6 infection, implying that the HHV-6 DNA detected in spermatozoa corresponds mainly to ciHHV-6. However, viral DNA may also be detected at a low level that is not in favour of the presence of ciHHV-6. Further studies are necessary to determine the origin of detected viral genomes.

The response of medical virology laboratories to the influenza A(H1N1)pdm09 outbreak in Paris Île-de-France region.

The outbreak of influenza A(H1N1)pdm09 was a challenge for the laboratories of Paris Île-de-France region in charge of virological diagnosis. In order to evaluate the quality of their response to this challenge, a retrospective survey based on a self-administered standardized questionnaire was undertaken among the 18 hospital laboratories involved in A(H1N1)pdm09 virus detection over a period of 10 months from April 2009 to January 2010. All concerned laboratories responded to the survey. Due to imposed initial biosafety constraints and indications, virological diagnosis was performed in only two laboratories at the start of the studied period. Step by step, it was further settled in the other laboratories starting from June to November 2009. From the beginning, A(H1N1)pdm09-specific RT-PCR was considered the reference method while the use of rapid influenza detection tests remained temporary and concerned a minority of these laboratories. Among the overall 21,656 specimens received, a positive diagnosis of influenza A(H1N1)pdm09 was obtained in 5,390 cases (25%), the positivity range being significantly higher among women as compared to men (P<0.0001) and subjects below 45 years of age as compared to those over 65 years (P<0.0001). Two peaks in positivity frequency were observed at weeks 24 (30%, 8-12 June 2009) and 44 (50%, 26-30 October 2009) respectively, the latter one occurring 2 weeks earlier than the peak of epidemic at the national level. In contrast, a low positivity rate was detected at weeks 38-40 in relationship with other respiratory virus infections which were clinically misinterpreted as a peak of influenza epidemic. These data demonstrate the ability of medical virology laboratories of Paris Île-de-France region to provide in real time a valuable diagnosis of A(H1N1)pdm09 virus infection and a relevant view of outbreak evolution, suggesting they will be a crucial component in the management of future influenza epidemics.

[Acute human herpesvirus 6 (HHV-6) infections: when and how to treat?]. / Infections aiguës à herpèsvirus humain 6 (HHV-6) : quand et comment traiter ?

The pathogenicity of human herpesvirus 6 (HHV-6) still raises numerous questions. Acute HHV-6 infections correspond to primary infections, reactivations or exogenous reinfections. The expression of related clinical symptoms is highly variable but may be extremely severe, particularly among immunocompromised patients. The prototypic severe disease associated with these infections is limbic encephalitis occurring in stem cell transplant recipients. The diagnosis of acute HHV-6 infections relies on the quantitation of viral load in whole blood by means of quantitative PCR. The demonstration of HHV-6 causative role in the genesis of clinical symptoms requires additional investigations such as the search for HHV-6 DNA in cerebrospinal fluid in case of encephalitis. The chromosomal integration of HHV-6 DNA is a rare event among HHV-6-infected subjects but may alter the interpretation of virological results. Therapy with ganciclovir, foscarnet or cidofovir has not yet clear indications but, at the current stage of knowledge, only concerns the treatment of highly symptomatic infections. The usefulness of prophylactic or pre-emptive antiviral chemotherapy has not yet been convincingly demonstrated. Treatment efficacy must be checked through a clinical and virological follow up, based in part on quantitative PCR approaches. Controlled studies are urgently needed with the goal of evaluating the cost-effectiveness of HHV-6 follow up and therapy in different clinical situations.

[Evaluation of the LightCycler 480 real-time PCR system for the measurement of CMV, EBV, HHV-6 and BKV viral loads in whole blood]. / Evaluation de la plate-forme de PCR en temps réel LightCycler 480 pour la mesure des charges virales CMV, EBV, HHV-6 et BKV dans le sang total.

OBJECTIVE: The Roche LightCycler 480 (LC480) system was evaluated for quantitative molecular diagnosis of opportunistic viral infections caused by human cytomegalovirus (CMV), Epstein-Barr virus (EBV), human herpesvirus 6 (HHV-6), and BK virus (BKV), in comparison with "in-house" real-time PCR assays. PATIENTS AND METHODS: A total of 253 whole blood specimens obtained from transplant recipients were tested. RESULTS: Both the "in-house" and the LC480 methods were highly correlated (Spearman correlation coefficient Rho> or =0.85; p<0.0001) with an excellent overall qualitative agreement (90.5%) and no significant quantitative difference between both techniques for the four viruses tested. The accuracy of the LC480 protocols were confirmed further by the results obtained with the 44 samples from the Quality Control for Molecular Diagnosis (QCMD) 2008 proficiency panel. CONCLUSION: The LC480 system constitutes a suitable and versatile real-time PCR platform in a routine laboratory setting for the diagnosis and monitoring of opportunistic viral infections in transplant recipients, by measuring HCMV, EBV, HHV-6, and BKV loads in whole blood samples.

Utilization of microsatellite polymorphism for differentiating herpes simplex virus type 1 strains.

The herpes simplex virus type 1 (HSV-1) genome is a linear double-stranded DNA of 152 kpb. It is divided into long and short regions of unique sequences termed U(L) and U(S), respectively, and these are flanked by regions of inverted internal and terminal repeats. Microsatellites are short tandem repeats of 1- to 6-nucleotide motifs; they are often highly variable and polymorphic within the genome, which raises the question of whether they may be used as molecular markers for the precise differentiation of HSV-1 strains. In this study, 79 different microsatellites (mono-, di-, and trinucleotide repeats) in the HSV-1 complete genome were identified by in silico analysis. Among those microsatellites, 45 were found to be distributed in intergenic or noncoding inverted repeat regions, while 34 were in open reading frames. Length polymorphism analysis of the PCR products was used to investigate a set of 12 distinct HSV-1 strains and allowed the identification of 23 polymorphic and 6 monomorphic microsatellites, including two polymorphic trinucleotide repeats (CGT and GGA) within the UL46 and US4 genes, respectively. A multiplex PCR method that amplified 10 polymorphic microsatellites was then developed for the rapid and accurate genetic characterization of HSV-1 strains. Each HSV-1 strain was characterized by its own microsatellite haplotype, which proved to be stable over time in cell culture. This relevant innovative tool was successfully applied both to confirm the close relationship between sequential HSV-1 isolates collected from patients with multiple recurrent infections and to investigate putative nosocomial infections.

Detection of human herpesviruses HHV-6, HHV-7 and HHV-8 in whole blood by real-time PCR using the new CMV, HHV-6, 7, 8 R-gene kit.

Human herpesviruses 6 (HHV-6), 7 (HHV-7) and 8 (HHV-8) are lymphotropic herpesviruses that may cause opportunistic diseases in immunosuppressed patients such as transplant or AIDS patients. The new commercial CMV HHV-6, 7, 8 R-gene kit (Argene, Varilhes, France) for the simultaneous quantitation of HHV-6 and qualitative detection of HHV-7 and HHV-8 was evaluated using whole blood samples (respectively, n=175, 100 and 161) and using different extraction and real-time PCR platforms in two Centers A and B. In comparison with HHV-6 in-house real-time PCR the commercial kit showed agreements of 96% (n=75) and 85% (n=100) in A and B, respectively, with significant Spearman's correlation between both techniques (in A: r=0.97 [p<0.001]; in B: r=0.70 [p<0.001]). The Bland-Altman test results and prospective monitoring of patients confirmed the accuracy of these HHV-6 real-time PCR techniques. The agreement between the in-house HHV-7 PCR and commercial kit was of 86% (n=100). In comparison with in-house HHV-8 real-time PCRs, the commercial kit showed agreements of 100% (n=61) and 93.7% (n=96) in A and B, respectively. These results demonstrate that the new commercial CMV HHV-6, 7, 8 R-gene kit was an efficient and reliable tool for the diagnosis of herpesvirus 6, 7, 8 infections.

Monitoring of human cytomegalovirus infection in immunosuppressed patients using real-time PCR on whole blood.

BACKGROUND: Quantitative monitoring of human cytomegalovirus (HCMV) is currently used in the follow-up of immunosuppressed patients. OBJECTIVE: To investigate whether real-time PCR quantification (QPCR) of HCMV DNA could replace pp65 antigenemia. STUDY DESIGN: We compared HCMV QPCR on whole blood (WB) and on plasma with a pp65-antigenemia assay on 192 samples. Afterwards, we tested 1310 samples from 308 immunosuppressed patients both by antigenemia assay and QPCR on WB. RESULTS: The first study comparison showed that QPCR results on WB and plasma were significantly correlated with antigenemia. QPCR on WB was more sensitive than QPCR on plasma or antigenemia, detecting 31 and 49 additional positive samples, respectively. During the second comparison, QPCR on WB and antigenemia were again correlated (r=0.70; p<0.0001), but QPCR detected 244 additional positive samples. HCMV DNA was detected earlier than pp65 antigen (median difference: 14 days; range: 7-30). One, 5, 10, 50 and 100 pp65-positive cells/200,000 leukocytes corresponded to 439, 1531, 2623, 9150 and 15,671 HCMV DNA copies/mL of WB, respectively, but this equivalence differed according to the sub-group of patients considered. CONCLUSION: QPCR on WB is the most sensitive method for the monitoring of HCMV infection in immunosuppressed patients.

Update on infections with human herpesviruses 6A, 6B, and 7.

Human herpesviruses 6A, 6B, and 7 (HHV-6A, HHV-6B, HHV-7) are genetically related to cytomegalovirus. They belong to the Roseolovirus genus and to the Betaherpesvirinae subfamily. They infect T cells, monocytes-macrophages, epithelial cells, and central nervous system cells. These viruses are ubiquitous and are responsible for lifelong chronic infections, most often asymptomatic, in the vast majority of the general adult population. HHV-6B is responsible for exanthema subitum, which is a benign disease of infants. HHV-6A and HHV-6B also cause opportunistic infections in immunocompromised patients: encephalitis, hepatitis, bone marrow suppression, colitis, and pneumonitis. Their etiological role in chronic diseases such as multiple sclerosis, cardiomyopathy, and thyroiditis is still controversial. The pathogenicity of HHV-7 is less clear and seems to be much more restricted. Chromosomal integration of HHV-6A and HHV-6B is transmissible from parents to offspring and observed in about 1% of the general population. This integration raises the question of potential associated diseases and can be a confounding factor for the diagnosis of active infections by both viruses. The diagnosis of HHV-6A, HHV-6B, and HHV-7 infections is rather based on gene amplification (PCR), which allows for the detection and quantification of the viral genome, than on serology, which is mainly indicated in case of primary infection. Ganciclovir, foscarnet, and cidofovir inhibit the replication of HHV-6A, HHV-6B, and HHV-7. Severe infections may thus be treated but these therapeutic indications are still poorly defined.

Genetic variability affects the detection of HIV by polymerase chain reaction.

Nine isolates of HIV-1 obtained from Congolese AIDS patients were amplified by the polymerase chain reaction (PCR) using primer pairs and oligomer probes derived from the HIV-1 LAV-BRU (BRU) sequence. When compared to BRU, two isolates exhibited a significant decrease of PCR efficiency with a given primer pair. Moreover, the DNA amplified from two other isolates did not hybridize with the corresponding probe despite efficient PCR. Base substitutions were detected in the regions of proviral genomes involved in oligonucleotide annealing and were assumed to be responsible for the failure of both amplification and probing. Our data confirm that the genetic variability of HIV-1 may reduce the efficiency of PCR as a diagnostic procedure, especially in the case of African isolates.

Multicentre quality control of polymerase chain reaction for detection of HIV DNA.

OBJECTIVE: Seven French laboratories tested the specificity and sensitivity of the polymerase chain reaction (PCR) for the detection of HIV-1 DNA. METHODS: Following its own PCR protocols, each laboratory independently tested blind two panels of 20 coded peripheral blood mononuclear cell samples collected from HIV-1-seropositive individuals and from HIV-1-seronegative individuals at high or low risk of HIV infection. For the first panel, laboratories were free to select type and number of primers; for the second, all were required to use the two primer pairs Pol 3/4 and MMy 9/10' (Nef 1). RESULTS: False-positive and false-negative results were observed in all laboratories (concordance with serology ranged from 40 to 100%). In addition, the number of positive PCR results did not differ significantly between high- and low-risk seronegatives. The use of crude cell lysates in DNA preparation produced the same PCR results as phenol-extracted DNA. Discrepancies between laboratories indicated that factors other than primer pairs contributed strongly to laboratory variability. CONCLUSIONS: Our results emphasize the importance of both positive and negative controls in PCR and demonstrate the value of multicentre PCR quality control.

Virological markers in the cerebrospinal fluid from HIV-1-infected individuals.

We analysed 127 specimens of cerebrospinal fluid (CSF) from 118 HIV-1-infected individuals at different stages of infection. Intrathecal antibody synthesis was evident in 23 samples tested and was more frequently directed against HIV than against rubella virus, herpes simplex virus, varicella zoster virus or cytomegalovirus. HIV was isolated from only 14% of the 127 CSF specimens, but from 82% of CSF-paired blood samples. HIV antigen was detected in 12% of CSF specimens and 44% of paired plasma samples. Twenty specimens analysed using the polymerase chain reaction (PCR) detected proviral DNA in 75% of CSF specimens. The low rate of virus recovery from CSF was caused by neither the freezing of specimens prior to culture nor therapy. In contrast, virus isolation from CSF was significantly associated with CSF cell count. Virus isolation and antigen detection in CSF were not correlated with either the Centers for Disease Control disease stage or the peripheral CD4+ lymphocyte count, whereas viraemia was significantly associated with a low CD4+ lymphocyte count. Moreover, virus isolation and antigen detection in CSF were not associated with symptoms of subacute HIV encephalitis, suggesting that these markers are not of potential value in the diagnosis of HIV-specific neurologic complications. The value of PCR in this field merits further investigation.

Thymocyte and thymic microenvironment alterations during a systemic HIV infection in a severe combined immunodeficient mouse model.

OBJECTIVE: A new model for systemic and multifocal HIV-1 infection was developed in severe combined immunodeficient (SCID) mice to study the alterations of thymocytes and of the thymic microenvironment that occur during a disseminated HIV infection. DESIGN AND METHODS: We grafted SCID mice with the classical human fetal thymus/liver co-implants together with fragments of autologous lungs (SCID-huLLT). These organs achieved normal differentiation and were productively infected after an intraperitoneal inoculation of two HIV-1 primary isolates. At time of sacrifice, thymic biopsies and thymic cell suspensions were analysed by immunohistochemistry, flow cytometry and lymphocyte function assays. RESULTS: At weeks 2-4 post-inoculation we observed the following thymocyte abnormalities: a minor to severe depletion of the immature CD1+CD4+CD8+ T cells (range, 0-73% thymocytes), compared with the persistence of mature CD4+ cells (11-50%) and amplification of CD8+ T cells (6-92%). The immature subset depletion was inversely related to the thymic HIV-1 viral load, suggesting the preferential infection of this subset. The residual mature thymocytes were functional as assessed by their sustained proliferative responses to CD3-triggering which contrasted with the lack of HIV-specific cytotoxic activity. A quantitative analysis of immunostained thymic sections revealed a disorganization and a densification of the thymic epithelial cells (TEC) network which occurred in all HIV-infected SCID-hu mice independently of the thymic CD1+CD4+CD8+ T-cell depletion. CONCLUSION: These results suggest that a systemic HIV infection induces in human thymuses from SCID-huLLT mice a preferential depletion of the immature thymocytes in the absence of mature CD4+ T-cell depletion, HIV-specific cytotoxic T-lymphocyte activity or thymic epithelial cell death, but is associated with dysplasia of the thymic microenvironment, and is therefore opening new perspectives for studying immune cell reconstitution strategies in HIV infection.

Comparative assessment of quantitative HIV viraemia assays.

OBJECTIVE: To compare two published methods for determining plasma viraemia in HIV-seropositive patients, with reference to a cellular viraemia assay. PATIENTS, PARTICIPANTS: Three patient groups were defined according to CD4 cell count: group I, less than 200 x 10(6)/l (23 patients); group II, 200-500 x 10(6)/l (18 patients); and group III, greater than 500 x 10(6)/l (13 patients). METHODS: The two reported methodologies were applied to all fresh samples, simultaneously and on the same day. RESULTS: The two techniques did not differ significantly in the detection of plasma viraemia: 82.3% of group I patients and 55% of group II patients were positive, while all group III patients were negative. Cellular viraemia was positive for 96% of the overall population. CONCLUSIONS: These results, obtained in a network of seven French laboratories involved in clinical trials, confirm that both plasma viraemia and cellular viraemia are useful virological markers.

Frequency of early in utero HIV-1 infection: a blind DNA polymerase chain reaction study on 100 fetal thymuses.

OBJECTIVE: To estimate the prevalence of in utero transmission of HIV-1 through the second trimester. MATERIAL AND METHODS: One hundred consecutive, unselected, intact fetuses, beyond 15 weeks gestational age (mean, 22.4 weeks) were studied. These were obtained following spontaneous intrauterine deaths (n = 4), miscarriages (n = 4), and elective mid-trimester terminations (n = 92), eight of which were fetuses with malformations from HIV-1-positive pregnancies. Coded DNA extracts from the fetal thymuses were tested blindly by polymerase chain reaction in three laboratories using a total of six different primer pairs. RESULTS: Two thymuses tested positive [95% confidence interval (Cl), 0.2-7]. Results from the three laboratories were consistent in all 100 cases. The two fetuses with HIV in the thymus both tested positive in other organs, demonstrating systemic HIV infection. The first fetus, whose mother had advanced AIDS, had died in utero and had diffuse toxoplasmosis. The second died following extremely premature delivery in a pregnancy complicated by repeated bleeding. HIV infection was observed in none of the 92 fetuses that resulted from elective mid-trimester terminations (95% Cl, 0-4). CONCLUSION: The frequency of early in utero HIV infection appears to be low, compared with transmission rates in infants born to HIV-1-infected mothers, suggesting that transmission occurs mostly later in pregnancy and/or at delivery. Specific risk factors may have implications in the occurrence of early as opposed to late transmission.

Determinants of paradoxical CD4 cell reconstitution after protease inhibitor-containing antiretroviral regimen.

OBJECTIVES: We evaluated the parameters influencing CD4 cell reconstitution after the introduction of highly active antiretroviral therapies in real life, as well as the frequency and the determinants of the discrepancies occurring between virus and CD4 cell count evolution. DESIGN AND METHODS: A total of 317 pre-treated patients starting a protease inhibitor (PI)-containing regimen were prospectively followed for 2 years on an intent-to-treat basis for CD4 cell counts and viral loads. RESULTS: The CD4 cell counts rapidly increased from baseline (50/mm3) by a median of 50/mm3 at month 2 (+0.72 CD4 cells/mm3/day) and up to 137/mm3 at the last follow-up (second slope: +0.16 CD4 cells/mm3/day). Two independent major factors among five parameters tested significantly affected the first phase, which was negatively correlated to the slope of CD4 cell decline before PI initiation, and was positively correlated to baseline CD4 cell counts (P = 0.0001); the second phase was mostly affected by the mean viral load reduction over time (P = 0.0001). Paradoxical CD4 cell reconstitution (15% of cases) was defined by a rapid or slow CD4 cell increase contrasting with a minor or strong viral reduction, respectively. The role of previous CD4 cell decline and the low effect of viral load reduction during the first 2 months explain the early paradoxical CD4 cell responses. The major influence of viral load reduction on the long-term reconstitution, however, reduces such paradoxical responses at 2 years. CONCLUSIONS: Early paradoxical CD4 cell reconstitution after the introduction of a PI are explained by the major influence of previous disease progression on the early CD4 cell increase, whereas the magnitude of viral load reduction over time reduces such paradoxical evolutions in the long term.

HIV-1 rebound during interruption of highly active antiretroviral therapy has no deleterious effect on reinitiated treatment. Comet Study Group.

BACKGROUND: Potent antiretroviral therapy (ART) with a protease inhibitor-based regimen is commonly used to treat HIV-1-infected patients. Transient treatment interruptions because of drug intolerance or other reasons are not uncommon. HIV-1 dynamics during therapy interruption and its consequences for the subsequent reinitiation of therapy have not been properly studied. METHODS: Ten antiretroviral-naive, HIV-1-infected subjects (mean baseline CD4 cell count of 414 cells/mm3 and plasma viral load of 4.8 log10 copies/ml) were treated with the triple drug ART regimen indinavir/zidovudine/lamivudine for 28 days. Therapy was then interrupted for 28 days, after which the same ART regimen was re-started. RESULTS: HIV-1 in plasma declined during the first 7 days of therapy with T1/2 of 1.5 days, and during days 7-28 with T1/2 of 8.9 days. Once therapy was interrupted, a delay of 4-7 days was observed in all subjects, preceding a rapid viral rebound with a mean doubling time of 1.6 days. Mean viral load after 28 days of interruption was 96% of baseline. Upon reinitiation of the same ART regimen, viral load declined at rates similar to those observed during the initial therapy (T1/2 of 1.6 and 8.0 days, respectively). No resistance-conferring mutations were observed in the HIV-1 reverse transcriptase (RT) and protease regions after the interruption of therapy. Plasma viral loads were maintained below 200 copies/ml in subjects continuing therapy for 4 (n = 9) to 12 (n = 5) months, with a mean CD4 cell count increase of 145 cells/mm3. CONCLUSIONS: The reintroduction of efficient ART therapy after a 1 month interruption shows viral kinetics similar to that of naive patients, and is not associated with the development of resistance. No deleterious effect on the reinitiated therapy was observed in patients who temporarily discontinued ART therapy. Nevertheless, because viral load rebounds back to baseline during treatment interruption, viral suppression is in effect put off by that period of time.

Re-occurrence of HIV-1 drug mutations after treatment re-initiation following interruption in patients with multiple treatment failure.

Antiretroviral treatment interruption in 20 extensively pre-treated HIV-1 patients with treatment failure led to genotype viral reversion of at least one class of drug-mutation resistance in half of the patients. The only predictive factor of reversion was found to be the duration of interruption. The outgrowth of residual wild-type virus seems not to be a true genetic reversion because drug mutations are detected rapidly at salvage therapy re-initiation.

Clinical and virologic characterization of acyclovir-resistant varicella-zoster viruses isolated from 11 patients with acquired immunodeficiency syndrome.

We studied the clinical resistance to acyclovir of infections with varicella-zoster viruses (VZV) in patients with acquired immunodeficiency syndrome, and we correlated it to virologic analyses. Eleven patients with VZV infections (treated with acyclovir, 30 mg/kg/day, given intravenously, or 4 g/day, given orally) were included in the study because of the failure of 10 days of acyclovir therapy. Susceptibility of VZV isolates to acyclovir was tested using a plaque reduction assay to determine the 50% inhibitory concentration (IC(50)) of acyclovir and the SI(50) (IC(50) of the patient isolate/IC(50) of the reference strain) to acyclovir. The thymidine kinase (TK) gene, which supports the resistance, was sequenced on amplified products. Only 3 patients had a significant increase in the IC(50), as compared with the IC(50) of the reference strain (SI(50) of > or =4), and a mutation in the TK gene. For the other 8 patients, the clinical resistance was not confirmed by the virologic results: the SI(50) was < 4, and no mutation was detected in the TK gene. Because no acyclovir-resistant strain appeared during a shorter period of time, we suggest an increase in the duration of the treatment to 21 days before acyclovir resistance is suspected.