点击显示 收起
1Laboratoire de Virologie, UPRES EA 2387, Groupe Hospitalier Pitié-Salpêtrière, and 2Laboratoire de Microbiologie, Faculté des Sciences Pharmaceutiques et Biologiques, Paris; 3Laboratoire de Virologie, UPRES EA 1156, Institut de Biologie des Hôpitaux de Nantes, and 4Service d'Hématologie Clinique, Hôtel Dieu, Nantes, France
Received 8 August 2002; revised 7 October 2002; electronically published 6 January 2003.
Human herpesvirus (HHV)6 and HHV7 loads were evaluated retrospectively in peripheral blood mononuclear cells (PBMC) from 78 recipients of stem cell transplantation (SCT) by real-time polymerase chain reaction. The median HHV-6 load in patients was 1357 genome equivalent copies (EqCop)/106 PBMC but was below the quantitation threshold in 31 immunocompetent individuals, which strongly suggests that HHV-6 reactivation occurred after SCT. The HHV-6 load was higher in patients with delayed neutrophil engraftment (P = .002) or severe graft-versus-host disease (P = .009). Moreover, the occurrence of at least 1 HHV-6related manifestation (fever, cutaneous rash, pneumonitis, or partial myelosuppression) was statistically associated with a concomitant virus load >103 EqCop/106 PBMC (P = .007). Conversely, HHV-7 reactivation was not favored, because median HHV-7 loads were similar in patients and healthy control subjects (1053 vs. 1216 EqCop/106 PBMC). The kinetics of Roseolovirus loads during the posttransplantation period suggested that HHV-7 may act as a cofactor of HHV-6 reactivation.
Financial support: Association pour la Recherche sur le Cancer (grant 5289); Association Claude Bernard.
Reprints or correspondence: Dr. Agnès Gautheret-Dejean, Laboratoire de Virologie, UPRES EA 2387, Groupe Hospitalier Pitié-Salpêtrière, 83 Boulevard de l'Hôpital, 75013 Paris, France ().
Human herpesvirus (HHV)6 and HHV7 are 2 closely related viruses classified into the Roseolovirus genus in the Betaherpesvirinae subfamily. A large majority of the general adult population is seropositive to both viruses [1, 2]. After primary infection, which commonly occurs in early childhood, they infect individuals persistently throughout life. CD4+ lymphocytes and epithelial cells of salivary glands appear to be the main cell targets of these viruses [3].
As for the other members of the Herpesviridae family, Roseolovirus reactivation or reinfection may be responsible for opportunistic diseases in immunocompromised hosts. In recipients of stem cell transplantation (SCT), HHV-6 infection has been associated with complications, such as encephalitis, interstitial pneumonitis, fever, cutaneous rash, delayed engraftment, transient myelosuppression, and high-grade graft-versus-host disease (GVHD) [35]. The pathological role of HHV-7 in patients undergoing SCT is still obscure. One case of HHV-7associated encephalitis has been reported in a blood SCT recipient [6]. The potential interactions between HHV-6 and HHV-7 after SCT have not been studied.
The classic tools of virological diagnosis, such as virus isolation in cell cultures or serological methods, are not adapted to the follow-up of Roseolovirus infections among large numbers of patients in a context of immunosuppression. The recent development of real-time polymerase chain reaction (PCR) methods, based on TaqMan technology, for HHV-6 and HHV-7 may constitute a useful tool for precisely quantifying virus loads in the peripheral blood mononuclear cells (PBMC) of patients, to assess the level of viral replication [7, 8].
Imbert-Marcille et al. [5] have recently studied the incidence and clinical relevance of HHV-6 infections in 92 unselected patients receiving allogeneic (allo) or autologous (auto) SCT, using qualitative DNA detection in PBMC and plasma by PCR. The aim of our retrospective study was to complete these data by use of real-time PCR methods, thereby allowing for the quantitation of Roseolovirus genomes. We have especially focused the analysis on the behaviors of HHV-6 and HHV-7 during the posttransplantation period and on the potential interactions between both viruses.
PATIENTS AND METHODS
Patients. Roseolovirus infections have been retrospectively studied in 78 patients undergoing SCT (26 allo-SCT and 52 auto-SCT recipients). All the patients were included in a larger prospective longitudinal study undertaken at Nantes University Hospital (Nantes, France) [5]. Patient characteristics related to sex, age, type of graft, source of stem cells, underlying disease, conditioning regimen, and antihuman cytomegalovirus (HCMV) prophylaxis are shown in . Median duration of posttransplantation follow-up was 107 days (range, 21350 days).
fig.ommitted
Table 1. Characteristics of stem cell transplantation (SCT) recipients.
The clinical features examined for their potential relationship with HHV-6 and HHV-7 were isolated fever (temperature, >38.5°C) for at least 2 days, with no positive microbiological findings documented in samples from any body site, acute GVHD diagnosed and scored according to organ injury and confirmed by performance of biopsy, cutaneous rash in the absence of confirmed GVHD, and pneumonitis with negative microbiology on bronchoalveolar lavage. No encephalitis was observed during the posttransplantation period in any patient. The following biological features also were examined: time to engraftment (an absolute neutrophil count >1 × 109 cells/L for 2 consecutive days within the study period), anemia (hemoglobin level, <85 g/L), leukopenia (leukocyte count, <3 × 109 cells/L), and thrombocytopenia (platelet count <5 × 1010 cells/L) after engraftment. The occurrence of clinical and biological events among patients is summarized in table 1.
Biological samples. A total of 685 sequential PBMC samples (261 for allo-SCT and 424 for auto-SCT recipients) were analyzed. A pretransplantation sample was available for 66 patients (20 allo-SCT and 46 auto-SCT recipients). Posttransplantation samples were collected either once a week or twice a month. Preparation of PBMC from blood samples has been described elsewhere [5]. Thirty-one PBMC specimens from healthy donors (21 men and 10 women; median age, 28 years) also were tested.
Viral DNA quantitation. DNA extraction from PBMC was performed by use of the High Pure Viral Nucleic Acid Kit (Roche Diagnostics), according to the manufacturer's instructions. DNA extracts were stored at -20°C until used. The quantitation of Roseolovirus DNAs was carried out in 10 L of DNA extract with the real-time PCR assays based on TaqMan technology that were developed recently in the laboratory [7, 8]. The number of cells present in the DNA extract was determined by use of a method that is based on the quantitation of the housekeeping gene of albumin, as described elsewhere [9]. Positive samples with a viral DNA extract result <10 genome equivalent copies (EqCop)/10 L were considered to be not quantifiable, because 10 EqCop/10 L was the inferior limit of the linear range of both methods, and the reproducibility of quantitation results under that threshold might be significantly affected. For samples with a viral DNA extract result >10 EqCop/10 L, the virus load was expressed as the number of viral EqCop per million of PBMC.
Statistical analyses. Data were analyzed by use of StatView 5.0 software. Frequencies were compared by use of Fisher's exact test. Statistical analyses concerning the comparison of quantitative values (i.e., virus loads and number of days) were assessed by the nonparametric Mann-Whitney U test, as appropriate. P < .05 was considered to be statistically significant.
RESULTS
Detection and Quantitation of Roseolovirus DNAs
Viral DNA detection. Before SCT, HHV-6 and HHV-7 DNAs were detected in 3 (4.5%) and 29 (43.9%) patients, respectively (). The frequency of HHV-6 DNA detection did not differ significantly from that observed for healthy donors (12.9%). By contrast, the HHV-7 DNA detection rate was significantly higher in healthy donors (87.0%) than in patients (P < .0001). During the posttransplantation period, HHV-6 and HHV-7 DNAs were detected at least once in 31 (39.7%) and 43 (55.1%) patients, respectively. The proportion of patients with HHV-6 DNA detection at any time after SCT was similar between allo- and auto-SCT recipients (30.8% vs. 44.2%; P = .328, not significant [NS]). The difference tended to be statistically significant regarding HHV-7 (38.5% vs. 63.5%; P = .053). No significant difference was evidenced according to sex, age, underlying disease, or conditioning regimen of patients. Among allo-SCT recipients, HCMV prophylaxis did not influence the rate of virus detection. Indeed, HHV-6 DNA was detected in 3 (25%) patients with high-dose acyclovir prophylaxis, compared with 5 (35.7%) patients without prophylaxis (NS). The corresponding rates for HHV-7 DNA detection were 6 (50%) and 4 (28.6%), respectively (NS). HHV-6 and HHV-7 genomes were simultaneously detected among 24 (30.8%) patients (5 [20.8%] allo-SCT recipients and 19 [79.2%] auto-SCT recipients). The concomitant detection of both viruses among SCT recipients during the posttransplantation period was significantly more frequent than that expected from a simple coincidence (P = .002, Fisher's exact test).
fig.ommitted
Table 2. Detection of Roseolovirus DNAs among stem cell transplantation (SCT) recipients and healthy donors.
Among the 619 PBMC samples from SCT recipients analyzed during the posttransplantation period, 115 (18.6%) were positive for HHV-6 DNA detection and 116 (18.7%) for HHV-7 DNA detection . The rate of HHV-6 DNA detection was significantly higher in samples from auto-SCT recipients (22.5%) than in samples from allo-SCT recipients (12.4%) (P = .002). The same difference was observed regarding HHV-7 DNA detection (24.1% vs. 10.4%; P < .0001). Forty samples (6.4%) exhibited a simultaneous detection of both viral genomes, which again indicated a nonrandom association between both infections (P < .0001, Fisher's exact test). With regard to the source of graft, HHV-7 DNA was detected in 94 (29.3%) samples from patients receiving peripheral blood stem cells and in 22 (7.3%) samples from patients receiving bone marrow stem cells (P < .0001). In contrast, the source of stem cells did not influence the rate of HHV-6 DNA detection: 57 (17.8%) samples positive for HHV-6 DNA after peripheral blood SCT versus 58 (18.1%) after bone marrow SCT (NS).
After SCT, the median time for the first virus detection was 19 days (range, 047 days ) for HHV-6 and 21.5 days (range, 0350 days) for HHV-7 (NS). Viral DNAs tended to be detected earlier in auto-SCT recipients than in allo-SCT-recipients (HHV-6, 18 vs. 27.5 days [P = .052]; HHV-7, 16.5 vs. 34 [P = .084]). According to the source of stem cells, HHV-7 DNA was detected on day 7 (range, 0350 days) among patients receiving peripheral blood stem cells and on day 34 (range, 14147 days) among patients receiving bone marrow stem cells (P = .0006). The kinetics of detection of Roseolovirus DNAs, expressed as the percentage of positive samples at different times of the study period, are shown in . For HHV-6, in agreement with the median detection time, a peak was observed at 34 weeks after SCT for all the patients, with a slightly earlier increase among auto-SCT recipients (week 2) than among allo-SCT recipients (week 3) (data not shown). The frequency of HHV-7 DNA detection, which was high before SCT, decreased during the first week after SCT and then increased again, tending to a pre-SCT percentage.
fig.ommitted
Figure 1. Kinetics of Roseolovirus DNA detection in patients who had undergone stem cell transplantation (SCT). Human herpesvirus (HHV)6 DNA () and HHV-7 DNA detection in peripheral blood mononuclear cells (PBMC) is represented before the graft (BG) and at different times after transplantation. Rate (%) of positive samples is indicated at each study time.
Viral DNA quantitation. HHV-6 load ranged from 188 to 1409 EqCop/106 PBMC (median, 640 EqCop/106 PBMC) in patients before the graft, whereas it was below quantitation threshold in healthy donors. In contrast, the HHV-7 load was similar among patients and healthy donors, ranging from 53 to 11,370 EqCop/106 PBMC (median, 541 EqCop/106 PBMC) and from 275 to 14,545 EqCop/106 PBMC (median, 1216 EqCop/106 PBMC), respectively (NS). During the posttransplantation period, the kinetics of virus loads showed a bimodal distribution, with the first period corresponding to an increase of virus loads, 100-fold, which was followed by a second period during which virus loads tended to return to values observed before the graft (). Concerning HHV-6, the first period lasted 4 weeks, with a peak at 2 weeks (median, 60,355 EqCop/106 PBMC), preceding the peak of detection rate (see ). For HHV-7, the first period lasted 2 weeks, with a peak at 1 week after SCT (median, 11,200 EqCop/106 PBMC) that also preceded the peak of detection rate. Regarding the samples from the entire posttransplantation period, the median HHV-6 load was 1357 EqCop/106 PBMC (range, 7117,272,727 EqCop/106 PBMC), and the median HHV-7 load was 1053 EqCop/106 PBMC (range, 15416,667 EqCop/106 PBMC). Thus, unlike HHV-6, the median HHV-7 load was similar among patients and healthy donors. With regard to the type of graft, the median HHV-6 load was significantly higher in allo- than in auto-SCT recipients (3580 vs. 958 EqCop/106 PBMC; P = .0004), whereas the median HHV-7 load was significantly higher in auto- than in allo-SCT recipients (1250 vs. 459 EqCop/106 PBMC; P = .032). However, the relevance of these differences must be considered with caution, because of the particular temporal distribution of virus loads (see ) and the wide variability of individual profiles among patients (see 3 examples of these profiles in a next section). Among allo-SCT recipients, Roseolovirus loads did not significantly differ regarding acyclovir prophylaxis (data not shown). No linear correlation between HHV-6 and HHV-7 loads was found when analyzing the 40 samples in which both genomes could be quantified concomitantly.
fig.ommitted
Figure 2. Evolution of Roseolovirus loads in patients who had undergone stem cell transplantation (SCT). Human herpesvirus (HHV)6 () and HHV-7 () median loads are represented (logarithmic scale) before the graft (BG) and at different times after transplantation.
Association with Clinical and Biological Features
The detection and quantitation of Roseolovirus DNAs were analyzed regarding clinical and biological features that occurred during the period after SCT. No significant association was evidenced with HHV-7 DNA detection. By contrast, the detection of HHV-6 DNA was statistically associated with cutaneous rash (P = .041) and thrombocytopenia (P = .018). Moreover, the occurrence of at least 1 clinical symptom (isolated fever, cutaneous rash, or pneumonitis) or biological event (anemia, leukopenia, or thrombocytopenia) was statistically associated with a concomitant HHV-6 load >103 EqCop/106 PBMC (P = .007), in particular partial myelosuppression (P = .017). In allo-SCT recipients, HHV-6 load was not correlated with the incidence of acute GVHD. Nevertheless, the HHV-6 load was significantly higher among the 6 recipients who developed a severe GVHD (grade IIIIV; 7971 EqCop/106 PBMC [range, 197419,149 EqCop/106 PBMC]) than among the 8 recipients with nonsevere GVHD (grade III; 1169 EqCop/106 PBMC [range, 713934 EqCop/106 PBMC]; P = .009). HHV-7 load tended to a similar difference (705 vs. 131 EqCop/106 PBMC; P = .064). Regarding neutrophil engraftment, the median HHV-6 load during the first month after transplantation was significantly higher among the 17 patients with a delayed engraftment (>20 days; 10,667 EqCop/106 PBMC [range, 322610,500,000 EqCop/106 PBMC]) than among the 61 patients without (2489 EqCop/106 PBMC [range, 117420,000 EqCop/106 PBMC]; P = .002). Furthermore, during the first 3 weeks after SCT, the proportion of samples with a HHV-6 load >104 EqCop/106 PBMC was statistically higher in patients with delayed engraftment (13.6%) than in others (4.5%; P = .036).
Profiles of Roseolovirus Loads and Virus-Associated Manifestations
The significant differences reported above were associated with a high individual variability, as illustrated by the great extent of virus load ranges. In this context, it is worth reporting the infection profiles of 3 particular SCT recipients who experienced clinical events possibly associated with HHV-6 reactivation .
fig.ommitted
Figure 3. Roseolovirus loads and clinical and biological time course of 3 patients who had undergone stem cell transplantation (SCT). Human herpesvirus (HHV)6 and HHV-7 loads are represented (logarithmic scale) for 2 allogeneic (allo)SCT recipients, patients 1 (A) and 2 (B), and for 1 autologous (auto)SCT recipient, patient 3 (C), before the graft (BG) and at different times after transplantation. Occurrence of clinical and biological events are above each graph (gray boxes). NQ, not quantifiable
Patient 1. Patient 1 was an allo-SCT recipient who exhibited a tremendous HHV-6 load increase during the posttransplantation period, whereas this parameter was not quantifiable before graft . During the initial increase of the HHV-6 load up to 107 EqCop/106 PBMC, the patient remained aplastic, with associated pyrexia and cutaneous rash. Transient leukopenia and thrombocytopenia then were observed between the third and the sixth weeks. The HHV-7 load remained at a low level throughout the whole follow-up period.
Patient 2. Two consecutive increases of HHV-6 load (up to 104 EqCop/106 PBMC) were observed for this allo-SCT recipient . During the first increase, concomitant with a peak in HHV-7 load (first month), the patient developed a transient thrombocytopenia, and biological signs of liver dysfunction were noted (data not shown). During the second elevation of virus load (third and fourth months), the patient was pyretic and developed anemia. In addition, this patient developed severe GVHD during the first and the third months.
Patient 3. In this auto-SCT recipient, an increase in HHV-7 load was observed during the first week after SCT . Parallel to the decrease in HHV-7 load, the HHV-6 load increased. Concomitantly, the patient, who had a delayed engraftment, developed a cutaneous rash and transient thrombocytopenia at the peak of HHV-6 load (104 EqCop/106 PBMC).
DISCUSSION
In the present study, Roseolovirus infections were studied retrospectively in 78 SCT recipients by using quantitation of viral genomes in PBMC by real-time PCR (TaqMan), which thus far has considered to be a good marker of viral replication. Concerning HHV-6, our results of DNA quantitation, using a qualitative detection of viral genome, confirmed those reported by Imbert-Marcille et al. [5], except for few discrepancies because of low HHV-6 loads. All HHV-6 strains were typed as variant B for patients [5] and healthy donors (data not shown).
Among healthy donors, the rate of HHV-6 DNA detection was low (12.9%), with a nonquantifiable load, whereas HHV-7 DNA was frequently detected (87.0%), with a higher level of replication (median load, 1216 EqCop/106 PBMC). This discrepancy, as reported elsewhere [10], may suggest that, in contrast to HHV-7, HHV-6 replication is controlled and suppressed in an immunocompetent context. Compared with healthy donors, the rate of HHV-6 DNA detection among patients before the graft (4.5%) was similar, whereas the rate of HHV-7 DNA detection (43.0%) was statistically lower. This might occur because of a specific impact of hematologic diseases or previous chemotherapy treatments on HHV-7 replication and would deserve additional studies.
In SCT recipients, HHV-6 and HHV-7 exhibited opposite behaviors during the posttransplantation period. In regards to HHV-6, the overall rate of DNA detection (39.7% of patients) was similar to previously reported results [11], and the level of viral replication (median virus load, 1357 EqCop/106PBMC) was statistically higher than in the control group of immunocompetent individuals (nonquantifiable load), which strongly suggests that HHV-6 reactivated in patients. Moreover, the virus reactivation seemed to occur during the first month, because a peak of DNA detection and a significant increase in virus load were observed during that time. Our data are in accordance with other reports [1113]. Furthermore, Yoshikawa et al. [14] found HHV-6 viremia 24 weeks after bone marrow transplantation using virus isolation, an accurate way to detect active HHV-6 infection. Those results indicate that HHV-6 reactivation occurs during the early post-SCT period, which is a critical phase for engraftment [15]. Data were different for HHV-7. Indeed, compared with healthy donors, median virus loads were similar (1053 vs. 1216 EqCop/106 PBMC), despite a very early peak in virus load (first week). Our findings are in accordance with previous studies that reported that the rate of HHV-7 DNA detection (57.0%) was similar before and after SCT [16] and that no elevation of HHV-7 load (using the semiquantitation method) occurred during the posttransplantation period [12]. These results suggest that, unlike HHV-6, HHV-7 does not reactivate after SCT.
Roseolovirus also had different behaviors according to the type of graft. As reported elsewhere [11, 17], the rate of viral DNA detection was not different after allo or auto graft in regards to the number of patients positive for detection at any time after transplantation. Nevertheless, in auto-SCT recipients, the rate of DNA-positive samples was higher for both viruses (HHV-6, 22.5% vs. 12.4%; HHV-7, 24.1% vs. 10.4%), and viral genomes tended to be detected earlier than in allo-SCT recipients. In contrast, HHV-6 median load posttransplantation was higher in allo-SCT recipients than in auto-SCT recipients, as illustrated by patient 1 who had undergone allo-SCT and exhibited a very strong HHV-6 reactivation during the posttransplantation period. On the contrary, the HHV-7 median load was higher in auto-SCT recipients than in allo-SCT recipients. These findings, similar to the results reported by Miyoshi et al. [17], suggest that immune mechanisms that control virus expression differ between HHV-6 and HHV-7.
HHV-7 DNA detection also was influenced by the source of stem cells. Indeed, in patients who received peripheral blood stem cells, the rate of DNA-positive samples was higher than in patients who received bone marrow stem cells (29.3% vs. 7.3%), and HHV-7 was detected earlier (7 vs. 34 days). Previous studies [18, 19] about immune reconstitution in SCT recipients showed that the time to engraftment is shorter in peripheral blood SCT recipients, compared with that in bone marrow SCT recipients (in this study: 10.5 vs. 19 days; P < .0001). The recovery of CD4+ T lymphocytes is notably accelerated. Thus, the faster recovery of the main target cell of HHV-7 could explain that viral DNA was detected in peripheral blood SCT recipients 3 weeks before bone-marrow SCT recipients. The larger cell tropism of HHV-6 may explain that the detection of this virus in PBMC was not influenced by the source of graft. However, our results are in contrast with those reported by Maeda et al. [12], who found that the rate of HHV-6 DNA detection was higher after bone marrow transplantation and that the rate of HHV-7 DNA detection was the same regardless of the source of stem cells. Further studies are needed to elucidate the influence of the source of graft on Roseolovirus infections in SCT recipients.
As reported elsewhere [11], Roseolovirus infections in SCT recipients were not influenced by either underlying disease or conditioning regimen. Moreover, high-dose acyclovir prophylaxis against HCMV infection did not significantly inhibit HHV-6 reactivation. These results confirm those of Chan et al. [11] but differ from those of Wang et al. [16]. Indeed, prophylactic treatment modalities of Roseolovirus infections (i.e., drugs and doses) in SCT recipients need to be clearly established.
No correlation between HHV-7 infection (DNA detection or quantitation) and clinical or biological manifestations was observed. Previous studies already reported no association with GVHD [11, 16]. As reported elsewhere by Wang et al. [16], we did not find any association with a delayed engraftment, which is in contrast with the results of Chan et al. [11]. Nevertheless, HHV-7 did not have any myelosuppressive effect on progenitor cells in vitro [20], which is consistent with the absence of viral hematopoietic effect in vivo. In contrast, HHV-6 could suppress proliferation and differentiation of progenitor cells in vitro [20, 21], and the amount of HHV-6 DNA in PBMC has been associated with a delayed platelet engraftment [12]. In the present study, thrombocytopenia was associated with HHV-6 DNA detection. Moreover, delayed neutrophil engraftment was associated with a statistically higher median HHV-6 load within the first month and with a higher proportion of positive samples with a virus load >104 EqCop/106 PBMC. Thus, in the follow-up of SCT recipients, the value of the HHV-6 load within the first weeks after transplantation should be considered as a prognostic factor with regard to the time to engraftment. If the threshold of 104 EqCop/106 PBMC is reached, the neutrophil recovery, which is potentially impaired, might be enhanced by granulocyte colony-stimulating factor treatment. The occurrence of a cutaneous rash also was associated with HHV-6 DNA detection, as reported in a study founded on virus isolation and DNA detection in skin biopsy specimens [22]. The relationship between HHV-6 infection and GVHD is still controversial. In our study, the severity of the acute GVHD was associated with HHV-6 load: the higher the grade of GVHD, the higher the load in PBMC. Wilborn et al. [23] found an association between HHV-6 DNA detection in PBMC and GVHD severity. Because of the high prevalence of Roseolovirus in the general population, the link with clinical or biological manifestations was difficult to show on the basis of a qualitative virus detection. In the present study, the analysis of HHV-6 loads in SCT recipients allowed us to define a threshold for the occurrence of HHV-6related manifestations. Indeed, virus loads >103 EqCop/106 PBMC were statistically associated with the occurrence of at least 1 of the clinical or biological events studied. This probably indicates that a high rate of viral replication is necessary for induction of clinical symptoms. The 3 examples of SCT recipients confirm the temporal relationship between HHV-6 (variant B) replication and clinical events, which also was illustrated by other reports of SCT recipients developing transient pyrexia, cutaneous rash, cytopenia, or liver dysfunction with either a concomitant HHV-6 DNA detection [24, 25] or a concomitant increase of HHV-6 load [2628].
Interactions between members of the Betaherpesvirinae subfamily have been increasingly studied in immunocompromised patients. In particular, HHV-7 has been described as a potential cofactor of HCMV in the progression to HCMV disease in kidney transplantation recipients [29] or SCT recipients [11]. In the present study, only 4 patients developed active HCMV infection, all without clear association with Roseolovirus infections (data not shown), but the question of interactions between HHV-6 and HHV-7 is pending. In vitro, another report has shown that HHV-6 could be reactivated from latency by infection with HHV-7 [30]. In vivo, clinical symptoms of exanthem subitum have been attributed to the reactivation of HHV-6 that followed primary HHV-7 infection [31]. To our knowledge, interactions among Roseolovirus in SCT recipients have not been studied yet. In our study, the simultaneous post-SCT DNA detection of HHV-6 and HHV-7 was significantly more frequent in patients than expected from single virus detection rates, but no correlation was found in term of virus loads. However, the kinetics of Roseolovirus loads in all SCT recipients, with a peak in HHV-7 load preceding the peak in HHV-6 load, as illustrated by the profiles of patients 2 and 3, may reflect the potential role of HHV-7 as a cofactor of HHV-6 replication and reactivation.
In summary, HHV-6 and HHV-7, 2 closely related virus members of the Betaherpesvirinae subfamily, exhibited distinct infection profiles in SCT recipients, which suggests different immune mechanisms regulating genome expression of each virus. As already described for others herpesviruses, HHV-6 reactivated during immunosuppression, whereas immunosuppression did not favor HHV-7 replication. The latter, which is present in almost all healthy individuals, appeared to be well adapted to human organism and did not induce any clinical or biological symptoms. Moreover, we saw a temporal relationship between HHV-6 replication and the occurrence of possible HHV-6related manifestations, and we defined a virus load threshold (103 EqCop/106 PBMC) over which HHV-6related manifestations were likely to occur. Therefore, the management of SCT recipients should include a sequential measurement of HHV-6 load to prevent viral complications, because specific antiviral therapy exists [32] and is in search of relevant indications.
References
1.
2.
3.
Dockrell DH and Paya CV. Human herpesvirus6 and 7 in transplantation. Rev Med Virol 2001; 11:2336.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.