HHV-6 and HHV-7 infections were documented in 15 patients each. Table 1 shows the compartments in which HHV-6 and/or HHV-7 were amplified. The most frequent sample analyzed was CSF, tested in 14/30 (46.6%) cases and 20/31 (64.5%) controls. Blood was used in 9/30 (30.0%) cases and 11/31 (35.5%) controls. HHV-6 and HHV-7 were also amplified in pericardial and nasopharyngeal specimens in seven of the 30 cases (23.3%).
Table 2 shows the clinical characteristics of patients in whom HHV-6 or HHV-7 were amplified. The most frequent symptomatology was neurological alterations (meningo-encephalitis, encephalopathies, coma, bradypsychia), which were present in approximately 80% for both viruses. Fever (42.9% in both viruses) and hepatitis (35.7% and 21.4%, respectively) were also frequent. Significantly, clinical enteritis was only present in patients with HHV-6 infection (40%) and none with HHV-7 infection.
Other viruses (not HHV-6 or HHV-7) were amplified in the previous 3 months (Table 3). Epstein Barr virus (EBV) was isolated in five cases, Enterovirus in two patients, and Cytomegalovirus (CMV) in one patient. On the other hand, in the control group, Epstein Barr virus was isolated in one patient and hepatitis B virus (HBV) in one patient. In 22 of the 30 cases, no new viral PCRs were performed in the following 3 months; but in the eight cases in which viral PCRs were repeated, HHV-6 was amplified in three patients (two of them in conjunction with CMV and EBV), CMV in one patient, and respiratory syncytial virus (RSV) in one patient. Another non-viral co-infection was confirmed in 16.1% of controls: four bacterial infections and one mycobacterial infection; however, this type of co-infection occurred in 48.3% of cases: seven bacterial infections and six fungal infections (p = 0.01). Non-viral co-infection was mostly in patients with HHV-6 infection (73.3% vs. 20% in HHV-7 infection). The frequency of non-viral infections in the following 3 months was also significant: 6.5% in controls and 33.3% in cases (p = 0.01).
There were eight hematologic diseases in the case group, five of them active. However, only one inactive disease was documented in the control group. Among cases, 31% were receiving corticosteroids (20.6%, with equivalent doses of prednisone ≥ 15 mg/d or had recently received them, 11%), while only one patient (3.2%) received corticosteroids, with doses < 15 mg/d of prednisone, in the control group. Previous or simultaneous chemotherapy was present in 20% of cases and none of the controls. There were also more treatments with biologic therapies among cases (23.3%) than among controls (6.5%). Thirteen percent of cases and none of the controls had had any episode of neutropenia (<500 cells/µl) in the last 3 months. The median lymphocyte count was also lower in cases (1416 lymph/µl) than in controls (2057 lymph/µl); 70.4% of cases had a lymphocyte count <1700 lymph/µl for only 41.4% of controls. There was 15% of hepatic injury (acute or chronic) in cases for none in controls. However, there was no difference between cases and controls in the development of renal failure (acute or chronic) or the control of the underlying disease. Finally, mortality was significantly higher in the case group (41.4%) than in the control group (13.3%)(p = 0.02).
Among risk factors associated with HHV-6 or HHV-7 infections, according to other reports, the following variables were included in multivariate analysis: hematologic disease, corticosteroid therapy, treatment with biological therapies or other immunosuppressive drugs, viral co-infection in the previous 3 months, and bacterial or fungal infection in the previous 3 months or the following 3 months (table 3). Finally, the variables that were independently associated with HHV-6 or HHV-7 infection were: corticosteroid therapy (OR: 24.1; 95%CI 2.6–223.1), the development of bacterial or fungal infections in the following 3 months (OR: 10.9; 95%CI: 1.9–62.0), and co-infection by other viruses in the previous 3 months (OR: 7.3; 95%CI: 1.2–44.8).
HHV-6 and/or HHV-7 infection was associated with a higher 1-year mortality (Tables 3 and 4). However, in multivariate analysis, factors independently associated with higher mortality were: age >60 years (OR 23.0; 95%CI: 1.9–269.8), no underlying disease control at 1 year (OR 17.2; 95%CI: 2.4–125.0), and bacterial or fungal infection in next 3 months (OR 14.6; 95%CI: 1.1–186.3).
When HHV-6 and HHV-7 infections were analyzed separately, differences in mortality were observed. An analysis that excluded HHV-7 infections from cases and included them among controls showed a higher 1-year-mortality in patients with HHV-6 infection (66.7%). A second multivariate model, including HHV-6 infection and adjusted by the independent variables confirmed in model 1 (age > 60 years, other infections and no underlying disease control at 1 year), confirmed that only age >60 years (OR 12.0; 95%CI: 2.0–69.4)(p: 0.005) and HHV-6 infection (OR 16.8; 95%CI 3.0–93.6) (p = 0.001) were independently associated with 1-year mortality. Although a higher rate of HHV6 infection was documented among patients with age >60 years (35.7% in >60 years vs. 18.2% in <60 years), no collinearity was detected (p = 0.15). On the other hand, HHV-7 infection alone was not associated with a higher 1-year mortality in uni- or multivariate analysis.
Finally, there was no homogeneous criterion regarding antiviral treatment. Among patients who presented HHV-6 infection (n = 15): seven did not receive any treatment, five received acyclovir, and three received ganciclovir/valganciclovir (two of them in combination with foscarnet). Among patients with HHV-7 infection (n = 15): ten received no treatment, three received acyclovir, and one received ganciclovir/valganciclovir and foscarnet in combination. In the control group, in 26/31, no treatment was administered. However, four received acyclovir and one received ganciclovir. There was no difference in 1-year outcome (death, renal or hepatic failure or underlying disease control) among treated and untreated patients.