|
|
| |||||||||||||||||||||||||||||||
Background Human herpesvirus 6 (HHV-6) is a recently described herpesvirus that is epidemiologically and biologically similar to cytomegalovirus. It is the cause of exanthem subitum (roseola) in children.
Methods To evaluate the possible role of HHV-6 infection in pneumonitis in immunocompromised patients, we used quantitative HHV-6 polymerase chain reactions to study lung-biopsy specimens from 15 patients with pneumonitis after bone marrow transplantation and lung tissue from 15 immunocompetent subjects without pneumonitis and 6 fetuses.
Results HHV-6 DNA was detected in lung tissue from all 15 patients, from 14 seropositive control subjects, and from none of the 7 seronegative control subjects. Six patients had levels of HHV-6 DNA in lung tissue that were 10 to 500 times higher than those in any of the other patients or control subjects. Increased levels of HHV-6 DNA correlated with a decreased risk of death from pneumonitis (P = 0.015), an increased severity of graft-versus-host disease (P = 0.023), and the presence of idiopathic pneumonitis (P = 0.037). Levels of HHV-6 DNA correlated directly with the changes in HHV-6 antibody titers in the interval between the pretransplantation period and the open-lung biopsy (P = 0.002). Low levels of HHV-6 antibody at the time of the open-lung biopsy were also associated with the diagnosis of idiopathic pneumonitis (P = 0.002).
Conclusions The concentrations of HHV-6 genome in lung tissue and their relation to changes in serologic titers support an association between HHV-6 infection and idiopathic pneumonitis in immunocompromised hosts.
HHV-6 shares many characteristics with cytomegalovirus, including DNA sequence homology, similar genomic organization, antigenic cross-reactivity, and similar in vitro growth characteristics18,19. Cytomegalovirus is commonly reactivated in immunosuppressed patients, causing a wide variety of syndromes including interstitial pneumonitis, which accounts for about 70 percent of the episodes of pneumonitis in marrow-transplant recipients20. But in this population about 20 percent of the episodes of pneumonitis cannot be attributed to any specific infectious agent even after rigorous laboratory investigation. The ubiquitous nature of HHV-6 and its close epidemiologic and genomic relation to cytomegalovirus led us to evaluate the association between HHV-6 and idiopathic pneumonitis in recipients of marrow transplants.
Methods
Patients and Tissue Sources
Fifteen marrow-transplant recipients at the Fred Hutchinson Cancer Research Center who had undergone open-lung biopsies for the diagnosis of pneumonitis were studied retrospectively. Patients with diverse clinical and histopathological diagnoses were selected, and only those with paired serum samples drawn before transplantation and at the time of open-lung biopsy were included. Fourteen of the 15 patients received their transplants between November 1986 and April 1988; 1 underwent transplantation in July 1983. The 15 patients had biopsies between February 1987 and August 1988 and made up 33 percent of the patients undergoing open-lung biopsy at the center during this time.
Lung-biopsy tissues from all 15 marrow-transplant recipients were extensively evaluated for infectious agents21,22. "Touch" imprints on slides, frozen sections, and permanent sections were stained according to standard techniques. Indirect fluorescent antibody analysis for cytomegalovirus and herpes simplex virus type 1 was performed on frozen sections from each sample with monoclonal antibodies 6-C5 (Genetic Systems, Seattle) and 3-G11 (Syva, Palo Alto, Calif.), respectively. A segment of fresh tissue from each biopsy was cultured for viruses, bacteria, fungi, and mycobacteria according to standard techniques23.
Control lung tissues were prospectively obtained from 15 immunocompetent subjects and 6 fetuses. Among the 15 immunocompetent subjects, 14 (mean age, 33 years; range, 0.25 to 54) were organ donors who died of injuries caused by a motor vehicle accident (4 subjects), myocardial infarction (3), sudden infant death syndrome (1), a gunshot wound (1), cardiac malformation (1), anaphylaxis (1), obstructive hydrocephalus (1), hanging (1), or head injury (1). One additional subject, who was 44 years old, had a partial lung resection for a pulmonary carcinoid tumor. Six frozen specimens of fetal lung tissue (gestational-age range, 58 to 124 days; mean, 82) composed another set of controls.
All tissues were randomly coded so that the scientists performing the polymerase chain reaction (PCR) procedure were unaware of the identity of any patient. An independent observer who had no knowledge of the PCR results retrospectively reviewed the charts of all 15 marrow-transplant recipients.
HHV-6 Serologic Analysis
An enzyme immunoassay was used to detect HHV-6 serum antibodies in a microtiter system with detergent extracts of cells infected with HHV-6 strain U1102 or mock-infected HSB-2 cells24. The end-point titer was taken as the reciprocal of the highest twofold dilution with an absolute absorbance value of more than 0.5 and a relative absorbance value that was at least twofold higher against viral antigen than the mock-infected cell antigen. All serum samples were run in duplicate, and sequential samples from an individual subject were included in the same analysis in a blinded comparison. The use of this enzyme immunoassay has been validated by comparing its results with those obtained by anticomplement immunofluorescence and HHV-6 Western blot assays25.
PCR Analysis
Sample preparations for PCR were performed in a dedicated "pre-PCR" room where operators wore protective clothing and took other precautions to prevent contamination of specimens26.
Five frozen sections (10 microm) of lung tissue were combined and digested overnight with 100 µg of proteinase K per milliliter. DNA was purified by extraction with phenol, phenol-chloroform, and chloroform and then precipitated with sodium acetate (300 mM), glycogen (100 µg per milliliter), and 1 ml of ethanol. The resulting purified DNA was resuspended in 0.2 ml of 10 mM TRIS (pH 8). Each specimen was amplified three times: once with HHV-6 primer pair 101R (genetic function unknown), once with HHV-6 primer pair 5R (exon homologous to that of the HCMV UL87 gene), and once with beta-globin primers. Both HHV-6 primer sets produced specific products when tested with DNA from 20 HHV-6 isolates, and both failed to react with purified DNA from 10 isolates of herpes simplex virus, 10 cytomegalovirus isolates, 1 strain of Epstein-Barr virus, 1 isolate of varicella-zoster virus, and 2 isolates of human herpesvirus 7. The reaction mixture contained 50 mM potassium chloride; 1.5 mM magnesium chloride; 10 mM TRIS-hydrochloric acid (pH 8.4); 1 U of recombinant Taq DNA polymerase (AmpliTaq, Perkin-Elmer Cetus); 200 microM each of deoxyguanosine triphosphate, deoxyadenosine triphosphate, deoxythymidine triphosphate, and deoxycytidine triphosphate (Pharmacia); 0.83 microM (2.5 x 1013 molecules) of each primer purified by high-performance liquid chromatography (Midland Certified Reagents, Midland, Tex.); and 10 microl of sample in a final volume of 50 microl. HHV-6 and beta-globin amplifications were performed at specimen dilutions ranging from 0 (undiluted) to 10-5. The thermocycling procedure consisted of denaturation at 94 °C for 6 minutes, 30 cycles of annealing at 55 °C for 1 minute, extension at 72 °C for 1 minute, and denaturation at 94 °C for 1 minute, followed by a final extension at 72 °C for 10 minutes. Parallel amplifications containing specimen DNA and 100 copies of exogenous HHV-6 DNA documented the absence of endogenous amplification inhibitors. The negative controls consisted of PCR reagents without sample DNA (one to three per run), and an aliquot of uninfected HSB-2 cells matched and processed with each specimen. None of the negative controls were positive on PCR during the study. Liquid hybridization was used to detect the amplification products by hybridizing 7 microl of the PCR products with 106 cpm of 32P-labeled probe and subjecting the hybrids to electrophoresis in a nondenaturing acrylamide gel27. Autoradiographs of the dried gels routinely produced specific bands when 10 HHV-6 genomes were added to the PCR.
Standard dilution curves of HHV-6 DNA and cellular DNA were included in every run (Figure 1). The band intensities of the amplified diluted specimens were compared with those of the standard dilution curves after amplification to estimate the number of HHV-6 and human genomes in each sample28. Values obtained by repeated measurements of the same specimens in different runs were within 0.6 log10 of each other. To compensate for variations in the sample size and the extent of DNA recovery after purification, the estimate of HHV-6 genomes was expressed as the number of HHV-6 genomes per million cellular genomes (HHV-6 genomes per 106 cells). For cases in which more than one tissue block was evaluated from a single patient, the final quantitation was derived by calculating the geometric mean of the values.
|
Spearman rank correlations were used to evaluate the correlates of DNA and antibody concentrations. The change in antibody titers between the pretransplantation period and open-lung biopsy was assessed with the Wilcoxon paired-sample test, and comparisons between groups were made with the Mann-Whitney U test29. All reported P values are based on two-tailed tests. Differences in the variability of HHV-6 DNA levels between the marrow-transplant recipients and the controls were assessed with the modified Levene test of Brown and Forsythe on log-transformed data30. This test compares the magnitudes of deviations from the median values in the two groups.
Results
HHV-6 PCR of Lung Tissues
HHV-6 DNA was detected in lung tissues from all 15 marrow-transplant recipients (Table 1) and all 14 seropositive controls; 1 seronegative control did not have detectable HHV-6 DNA in his lung tissue, and none of six specimens of fetal lung tissue contained HHV-6 DNA (Table 2). Twenty-eight of the 30 blocks of frozen tissue from the 15 marrow-transplant recipients were positive for HHV-6 on PCR. Tissue sections from all 30 patient blocks were evaluated by both the 5R and 101R HHV-6 primer sets, which amplify different regions of the viral genome; the results were similar in all specimens.
|
|
The potential of HHV-6 DNA carryover during the process of cutting the samples with a microtome was ruled out in two ways. First, when the cutting order was compared with the levels of HHV-6 DNA in each specimen, there was no indication that specimens positive for HHV-6 DNA were more likely than negative specimens to follow samples with high levels of HHV-6 DNA. Second, negative samples remained negative on PCR when specimens containing high levels of HHV-6 DNA were alternately sectioned with blocks containing no HHV-6 DNA.
HHV-6 Serologic Analysis
Serum HHV-6 antibody titers were determined in samples obtained before transplantation and near the time of the open-lung biopsy (median, 1 day after biopsy; range, 9 days before to 11 days after biopsy).
All patients were seropositive for HHV-6 before transplantation, with a median antibody titer of 1600, whereas the median titer after transplantation at open-lung biopsy was 800 (P = 0.224 by the Wilcoxon paired-sample test). Four patients had fourfold or greater increases in their HHV-6 antibody titers, six patients had decreases of that magnitude, and five patients had little or no change in titers (Table 1). All six of the patients with high levels of HHV-6 DNA in their lung-biopsy specimens had a change in serum HHV-6 antibody titers of at least 200 percent. The magnitude of the change in HHV-6 antibody titers was positively correlated with the level of HHV-6 DNA (P = 0.002 by Spearman rank correlation). Finally, antibody titers at the time of open-lung biopsy were significantly lower in patients with idiopathic pneumonitis than in the other patients (P = 0.002 by the Mann-Whitney U test). No correlation was observed between the levels of or changes in cytomegalovirus antibody titers and the levels of HHV-6 DNA (Table 1).
Clinical and Histologic Features
Fourteen patients (eight female and six male) received allogeneic marrow transplants, and one patient (Patient 15, male) received an autologous marrow transplant. Their median age was 30 years (range, 3 to 45) (Table 1). The onset of pulmonary symptoms occurred a median of 48 days after transplantation (range, 9 to 1676), and open-lung biopsy was performed a median of 13 days after the onset of symptomatic pneumonitis (range, 1 to 93) (Table 1).
A retrospective, blinded chart review determined the maximal severity of graft-versus-host disease (GVHD) between the time of marrow infusion and open-lung biopsy. The GVHD scores (a score of 0 indicated no GVHD, and a score of 4 maximal disease) correlated with the levels of HHV-6 DNA (P = 0.023, Spearman rank correlation) (Figure 2). The levels of HHV-6 DNA were not associated with age, sex (Table 1), the pretransplantation conditioning regimen (cyclophosphamide [Cytoxan] and total-body irradiation, Patients 1 through 7 and 9 through 14; cyclophosphamide and busulfan, Patient 15; or cyclophosphamide, busulfan, and total-body irradiation, Patient 8), or the pretransplantation diagnosis (chronic myelogenous leukemia in Patients 1, 2, 3, 4, 12, and 14; acute lymphocytic leukemia in Patients 5, 9, and 13; lymphoma in Patients 8 and 10; acute nonlymphocytic leukemia in Patient 6; acute erythroid leukemia in Patient 11; acute myelogenous leukemia in Patient 3; and Hodgkin's disease in Patient 15). In general, the levels of HHV-6 DNA were higher in those who survived the episode of pneumonitis than in those who died (P = 0.015 by the Mann-Whitney U test) (Table 1). Other data from the retrospective chart review did not reveal significant differences between the group with high levels of HHV-6 DNA and the group with low levels with respect to the administration of antiviral or immunosuppressive drugs, the pretransplantation conditioning regimen, the history of blood-product therapy, or interpretations of chest radiographs.
|
Discussion
Quantitative PCR analysis revealed that 6 of the 15 marrow-transplant recipients who were studied had substantially higher levels of HHV-6 DNA in their lung tissues than the other transplant recipients or the controls. The levels of HHV-6 DNA were elevated mainly in patients with pneumonitis for which thorough analyses failed to identify fungi, bacteria, or other viruses as the cause of the illness. Five of the eight patients with idiopathic pneumonitis had high levels of viral DNA, as compared with only one of seven of those with pneumonitis with identified causes.
The specificity of our PCR assay for HHV-6 was supported by the concordance of the results with the use of two primer pairs from different regions of the genome. We did not find HHV-6 DNA in the one seronegative control or in the six samples of fetal lung tissue. We did, however, detect HHV-6 in normal lung tissue from all 14 seropositive adult controls. These data led us to develop a quantitative measurement to distinguish high levels of HHV-6 DNA, which may be more consistent with the occurrence of reactivation, from the lower levels found in normal adults. The quantitative differences in the viral DNA concentration among the subjects were reproducibly demonstrated in several separate experiments that used coded specimens. Furthermore, multiple coded samples amplified from the same subject consistently supported the classification of that subject into either the group with high levels of HHV-6 DNA or the group with low levels.
The HHV-6 antibody titers alone did not predict which patients had high levels of HHV-6 DNA in their lungs. However, all six patients with high levels of HHV-6 DNA had significant changes in HHV-6 antibody titers from the samples obtained before transplantation to those obtained afterward. Six of the 10 patients with substantial changes in HHV-6 antibody titers had decreases in the titers rather than increases, a trend that has also been described with other herpesviruses in marrow-transplant recipients31. Decreases in herpesvirus antibody titers may reflect general decreases in immune response associated with the transplantation procedure or the formation of immune complexes between HHV-6 antibodies and antigens, which could cause a decrease in free antibody levels. The multiple transfusions received by this patient population also confound serologic interpretations.
Surprisingly, only one of six patients (17 percent) with high levels of HHV-6 DNA died of pneumonitis during the episode studied, whereas six of nine patients with low levels of HHV-6 DNA died of pneumonitis. One explanation for this observation is that HHV-6-associated pneumonitis is not as severe as pneumonitis caused by cytomegalovirus or other agents in this selected patient population. However, the apparent difference in outcome could also be due to numerous other factors, such as the effects of therapy or differences in clinical management. Prospective studies are therefore needed to confirm this observation.
The severity of GVHD correlated with pulmonary levels of HHV-6 DNA. This observation is consistent with previous reports that correlated the severity of GVHD with the incidence of idiopathic pneumonitis32. Three possible explanations for these data are that the reactivation of HHV-6 exacerbates GVHD as well as causes pneumonitis; GVHD induces the reactivation of HHV-6, which in turn produces pneumonitis; and GVHD induces the reactivation of HHV-6, but the reactivation is unrelated to the occurrence of pneumonitis. Regarding the third possibility, we do not know whether high levels of HHV-6 DNA also occur in lung tissue from marrow-transplant recipients without pneumonitis because we could not perform open-lung biopsies on patients without pneumonitis. Similarly, biopsy tissues obtained from other organs at the time of pneumonitis were not available, precluding PCR analyses for HHV-6 at other sites. PCR analysis of tissues obtained at autopsy (spleen, adrenal gland, kidney, and lung) from three other marrow-transplant recipients revealed levels of HHV-6 DNA below those found in the patients with putative HHV-6-related pneumonitis described here. More detailed prospective studies are needed to evaluate whether the reactivation of HHV-6 is local or systemic.
In summary, marrow-transplant recipients with high levels of HHV-6 DNA in lung tissue tended to have idiopathic pneumonitis, a favorable outcome, severe GVHD, and changes in HHV-6 antibody titers, suggesting an association between HHV-6 infection and idiopathic pneumonitis. Prospective studies are needed to define the pathogenic potential of HHV-6 in relation to idiopathic pneumonitis in recipients of marrow transplants.
Supported by grants from the National Institutes of Health (1 R29 AI30648-01) and the National Cancer Institute (CA 15704, CA 18029, and CA 47748).
We are indebted to the late Dr. Robert Honess, whose willingness to share his scientific insights contributed much to our work; to Dr. Phil Pellet for the HHV-6 clones; to Drs. Zwi Berneman and Niza Frenkel for human herpesvirus 7 DNA; to Helen Newman-Gage and Ted Wrigley of the Northwest Tissue Center for control lung tissues; and to Clara Bryan, Patti Wilson, and Karin Rogers for technical assistance.
Source Information
From the Departments of Laboratory Medicine (R.W.C., M.-L.W.H., M.M., R.A., L.C.), Medicine (L.C.), Pathology (R.C.H.), Pediatrics (R.A.B.), and Statistics (J.Z.), University of Washington, and the Departments of Pathology (R.C.H.) and Infectious Diseases (R.A.B., J.D.M.), Fred Hutchinson Cancer Research Center, both in Seattle. A portion of this work was presented in abstract form at the 14th International Herpesvirus Workshop, Nyborg Strand, Denmark, August 20-26, 1989.Dr. Joel D. Meyers is deceased.
Address reprint requests to Dr. Cone at Children's Hospital and Medical Center, CH-82, 4800 Sand Point Way NE, Seattle, WA 98105.
References
| |||||||||||||||||||||||||||||||
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | TERMS OF USE | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2009 Massachusetts Medical Society. All rights reserved. |
Previous
