- Journal List
- HHS Author Manuscripts
- PMC3439599
As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsem*nt of, or agreement with, the contents by NLM or the National Institutes of Health.
Learn more: PMC Disclaimer | PMC Copyright Notice
Biol Blood Marrow Transplant. Author manuscript; available in PMC 2013 Nov 1.
Published in final edited form as:
Biol Blood Marrow Transplant. 2012 Nov; 18(11): 1700–1708.
Published online 2012 May 26. doi:10.1016/j.bbmt.2012.05.012
PMCID: PMC3439599
NIHMSID: NIHMS380924
PMID: 22641196
Danielle M Zerr, MD, MPH,1,2 Michael Boeckh, MD,3,4 Colleen Delaney, MD, MSc,1,4 Paul J Martin, MD,3,4 Hu Xie, MS,4 Amanda L Adler, BA,2 Meei-Li Huang, PhD,4 Lawrence Corey, MD,4 and Wendy M Leisenring, ScD4
Author information Copyright and License information PMC Disclaimer
The publisher's final edited version of this article is available free at Biol Blood Marrow Transplant
Abstract
Background
Human herpesvirus 6 (HHV-6) reactivation has been associated with acute graft-versus-host-disease (aGVHD), cytomegalovirus (CMV) reactivation, and mortality after allogeneic hematopoietic cell transplantation (HCT), but previous studies have yielded inconsistent results. We performed a large prospective study of allogeneic HCT recipients in order to more definitively define the relationships between HHV-6 and these important outcomes.
Methods
Plasma specimens were collected prospectively from 315 allogeneic HCT recipients and tested for HHV-6 DNA at baseline and twice-weekly for 12 weeks. Cox proportional hazards models were used to evaluate the time-dependent associations between HHV-6 reactivation and the targeted outcomes.
Results
HHV-6 was detected in 111 (35%) of 315 patients at a median of 20 days after HCT. HHV-6 reactivation was associated with subsequent CMV reactivation [adjusted hazard ratio (aHR) 1.9, 95% confidence interval (CI) 1.3-2.8, p=0.002]. High-level HHV-6 (>1,000 HHV-6 DNA copies/mL) was associated with subsequent grades II-IV aGVHD (aHR 2.4, 95% CI 1.60-3.6), p<0.001). High-level HHV-6 reactivation was also associated with non-relapse mortality (aHR 2.7, 95% CI 1.2-6.3, p=0.02).
Conclusion
HHV-6 reactivation was independently and quantitatively associated with increased risk of subsequent CMV reactivation, aGVHD, and mortality after HCT. A randomized antiviral trial is warranted to establish causality between HHV-6 and these endpoints and to determine if reducing HHV-6 reactivation will improve outcome after HCT.
Introduction
Human herpesvirus 6 (HHV-6) infects 90-100% of individuals during early childhood.[1] After primary infection, HHV-6 establishes a latent infection in hematopoietic reservoirs. This latent infection can become active in settings of severe immunosuppression, especially hematopoietic cell transplantation (HCT). Prior studies show that HHV-6 reactivates in approximately 40% of patients after HCT[2-4] and reactivation has variably been associated with important outcomes, including cytomegalovirus (CMV) reactivation,[5] acute graft-versus-host-disease (aGVHD),[2,6,7] and increased mortality.[4,6] Whether HHV-6 is actually causally associated with these problems remains controversial.
We performed a large prospective study of HHV-6 in HCT recipients in an effort to better understand the relationships between HHV-6, CMV, aGVHD, and mortality in these patients.
Material and Methods
The data presented in this report were generated in part from a prospective study designed to evaluate the associations between HHV-6 reactivation and neuropsychiatric and neurocognitive outcomes.[8] The protocol was approved by the Fred Hutchinson Cancer Research Center Institutional Review Board.
Subjects
Patients of all ages undergoing allogeneic HCT from January 2005 through August 2008 were eligible for enrollment. Those with limited English proficiency were excluded due to the frequent neuropsychiatric and neurocognitive assessments required for the parent study. The study was presented to 880 patients during the pre-transplantation evaluation and 474 were preliminarily enrolled (Figure 1). Of the 474, a total of 152 withdrew or were deemed ineligible before contributing data, leaving 322 patients who contributed data. Six patients with HHV-6 DNA levels suggestive of chromosomal integration[9] (determined a priori as increasing HHV-6 plasma DNA levels during the first 2 weeks after HCT and persistent levels ≥ 100 copies per/mL in ≥ 80% of subsequent plasma samples) were excluded from analyses. An additional patient, who contributed only baseline data, was also excluded. Of the 315 included patients, nearly all (n = 308, 98%) were followed for ≥ 4 weeks after HCT or until death.
Figure 1
Consort diagram
Clinical care
Participation in this study had no impact on clinical decisions, including those involving conditioning regimens, type of transplantation, aGVHD prophylaxis and treatment, or administration of antimicrobials. There were no recipients of T cell depleted stem cell grafts. CMV reactivation was monitored and treated per clinical standards of care. From the initiation of the study through February 2007, the primary mode of CMV screening was CMV antigenemia. After this point, plasma CMV PCR became the primary means of CMV screening. Patients were tested weekly for evidence of CMV reactivation through approximately day 100 following HCT. A pre-emptive antiviral therapy approach was followed.[10,11] Ganciclovir was the first-line antiviral post-engraftment, and foscarnet was the second.
Study procedures
Baseline (pre-HCT) and twice-weekly plasma specimens were collected through day 84 post-HCT for HHV-6 testing. A total of 6,255 specimens were obtained (85% of planned). Patients were followed through day 200 for mortality.
Clinical data and definitions
Demographic, clinical, and laboratory information was collected from clinical records and databases.
Underlying disease was categorized as “less advanced” or “more advanced” (Table 1).[12]
Table 1
Demographic and clinical characteristics of the cohort, overall and stratified by ever having HHV-6 reactivation.
| Overall N=315 n (%) | HHV-6 Reactivation N=315 | ||
|---|---|---|---|
| No n=204 | Yes n=111 | ||
| Age | |||
| • <=20 yrs | 63 (20%) | 33 (16%) | 30 (27%) |
| • 21-30 yrs | 22 (7%) | 12 (6%) | 10 (9%) |
| • 31-40 yrs | 31 (10%) | 22 (11%) | 9 (8%) |
| • 41-50 yrs | 51 (16%) | 34 (17%) | 17 (15%) |
| • >50 yrs | 148 (47%) | 103 (50%) | 45 (41%) |
| Female Sex | 126 (40%) | 82 (40%) | 44 (40%) |
| Race/Ethnicity | |||
| • Caucasian | 257 (82%) | 172 (84%) | 85 (76%) |
| • Hispanic | 10 (3%) | 6 (3%) | 4 (4%) |
| • Native American | 6 (2%) | 3 (2%) | 3 (3%) |
| • Other | 42 (13%) | 23 (11%) | 19 (17%) |
| CMV-donor/recipient serostatus | |||
| • Donor+/Recipient+ | 80 (25%) | 48 (24%) | 32 (29%) |
| • Donor+/Recipient- | 31 (10%) | 21 (10%) | 10 (9%) |
| • Donor-/Recipient+ | 92 (29%) | 57 (28%) | 35 (31%) |
| • Donor-/Recipient- | 112 (36%) | 78 (38%) | 34 (31%) |
| Positive HSV serostatus | 265 (85%) | 167 (83%) | 98 (88%) |
| Medical co-morbidity | |||
| • Low | 93 (30%) | 52 (26%) | 41 (37%) |
| • Moderate | 95 (30%) | 68 (33%) | 27 (24%) |
| • High | 127 (40%) | 84 (41%) | 43 (39%) |
| More advanced underlying diseasea | 174 (55%) | 109 (53%) | 65 (59%) |
| Total body irradiation | |||
| • ≥1200 cGy | 72 (23%) | 36 (17%) | 36 (33%) |
| • ≤400 cGy | 127 (40%) | 81 (40%) | 46 (41%) |
| • None | 116 (37%) | 87 (43%) | 29 (26%) |
| Conditioning regimen | |||
| • Myeloablativeb | 169 (54%) | 110 (54%) | 59 (53%) |
| • Nonmyeloablativec | 112 (35%) | 79 (39%) | 33 (30%) |
| • Reduced intensityd | 34 (11%) | 15 (7%) | 19 (17%) |
| HLA match | |||
| • Matched related | 98 (31%) | 74 (36%) | 24 (22%) |
| • Matched unrelated | 142 (45%) | 95 (47%) | 47 (42%) |
| • Mismatched relatede | 10 (3%) | 4 (2%) | 6 (5%) |
| • Mismatched unrelatedf | 65 (21%) | 31 (15%) | 34 (31%) |
| Stem cell source | |||
| • Bone marrow | 61 (19%) | 34 (17%) | 27 (24%) |
| • Cord blood | 21 (7%) | 6 (3%) | 15 (14%) |
| • Peripheral blood stem cell | 233 (74%) | 164 (80%) | 69 (62%) |
Open in a separate window
CMV = cytomegalovirus, HSV = herpes simplex virus, HLA = human leukocyte antigen
a“more advanced” underlying disease refers to diagnoses other than acute myeloid leukemia, acute lymphoblastic leukemia, or lymphoma in 1st remission, chronic myeloid leukemia in chronic phase, and refractory anemia without excess blasts.
bMyeloablative regimens included: any regimen containing ≥800 cGY total body irradiation, any regimen containing carmustine/etoposide/cytarabine/melphalan (BEAM), or any regimen containing busulfan/cyclophosphamide with or without antithymocyte globulin. The most common regimens included busulfan and cyclophosphamide (n=73), cyclophosphamide and ≥1200 cGY TBI (n=47), and cyclophosphamide,fludarabine, and 1320 cGY TBI (n=13).
cThe non-myeloablative regimen was fludarabine 90mg/m2 +/- TBI ≤300 cGY. The most common regimens included Fludarabine and 200 cGY TBI (n=76) and fludarabine and 300 cGy TBI (n=9).
dAll other regimens were considered reduced intensity. The most common regimen was treosulfan and fludarabine (n=17).
eAllele or antigen mismatch: 9/10 (n=4), 6/10 (n=1), 5/10 (n=5)
fAllele or antigen mismatch: 9/10 (n=40), 8/10 (n=4), cord blood transplant recipients (n=21) - all cord blood transplants were mismatched and a subset (n=16) were double cord blood transplants
Medical co-morbidity was defined and categorized using a validated scale.[13]
Pre-transplant lymphopenia was assessed at the last lymphocyte count obtained prior to starting conditioning chemotherapy and was defined as a lymphocyte count <600 (approximate lowest quartile).
Conditioning regimens were categorized as “myeloablative”, “nonmyeloablative” or “reduced intensity”. A variety of cytoreductive regimens were used; the most common regimens are reported in Table 1 by myeloablative category.
HHV-6 reactivation was defined as any level of plasma HHV-6 DNA.
High-level HHV-6 reactivation was defined as ≥ 1000 HHV-6 DNA copies/mL plasma. This level was chosen because it is a threshold for CMV that is commonly used to initiate use of pre-emptive antiviral therapy. In addition, in the context of this study, it is close to the median maximum level (873 HHV-6 copies/mL plasma).
CMV reactivation was defined as any level of plasma CMV DNA or whole blood antigenemia.
High-level CMV reactivation was defined as ≥ 1,000 CMV DNA copies/mL plasma or 10 cells/high-powered field of CMV antigenemia.
Acute graft versus host disease (aGVHD) grades and organ-specific types (skin, gastrointestinal, and liver) and stages were categorized as previously described by a single investigator (PJM) blinded to HHV-6 study results.[14]
Chronic graft versus host disease (cGVHD) was categorized as previously described.[15]
Overall mortality was defined as mortality occurring for any reason.
Non-relapse mortality was defined as mortality occurring for reasons other than relapse in patients receiving myeloablative HCT or for reasons other than relapse or progression of underlying disease in patients receiving non-myeloablative HCT.
Antivirals that might effect CMV or HHV-6 activity were categorized as “low activity” (acyclovir) or “high activity” (foscarnet, ganciclovir, or cidofovir).
Laboratory procedures
Individuals blinded to patients’ clinical status performed PCR analyses as previously described.[8] Detection of 1 copy of HHV-6 DNA/reaction (25 copies/ml of plasma) was considered a positive specimen. All HHV-6 DNA identified by PCR was typed as HHV-6 A or B.[16]
Statistical Analyses
Cox proportional hazards models were utilized to evaluate the impact of HHV-6 on the hazards of subsequent occurrence of each of the endpoints of interest for this study: CMV reactivation, aGVHD, cGVHD, and mortality. Mortality was assessed through day 200 while CMV, aGVHD was assessed through 100 days, and cGVHD endpoints were assessed both through day 100 and 1 year. Both grades II-IV and III-IV overall aGVHD were evaluated. In addition, organ-specific aGVHD was examined using the stages that specifically inform overall aGVHD grades II-IV and III-IV: stages 3-4 and 4 skin subtypes and stages 1-4 and 2-4 liver and gastrointestinal subtypes. The number of observations was insufficient to perform multivariable analyses of grade IV overall aGVHD. We also evaluated a composite endpoint of any mortality, grade II-IV aGVHD, or any level of CMV reactivation to evaluate the impact HHV-6 reactivation has on the likelihood of remaining alive and free of aGVHD and CMV reactivation through 200 days following HCT. The key risk factor of interest for all analyses was HHV-6 reactivation, modeled as a time-dependent covariate and coded as positive using 2 definitions (modeled separately): 1) any level of detectable HHV-6 and 2) HHV-6 ≥1000 copies/mL plasma. Additional covariates included baseline demographic and clinical variables (Table 1). When HLA match was analyzed as a covariate, three strata were used: “matched (10 of 10 allele and antigen matched) related” versus “matched unrelated” versus “mismatched related” plus “mismatched unrelated”. Mismatched related and mismatched unrelated were combined into one category due to the small number of mismatched related cases (n=10). We also evaluated pre-transplant lymphopenia as defined above in each of the models. In addition, the dose of CD34-postive cells, use of antithymocyte globulin, female donor/male recipient status, and prior HCT were investigated in models for aGVHD. Administration of antiviral medications was also examined as a covariate for CMV reactivation. As an initial variable selection step, each factor was evaluated in a bivariable model with HHV-6 and considered eligible for the multivariable model if the two-sided p-value was <0.20 or if its inclusion modified the effect of HHV-6 by >10%. Once included in a full multivariable model, in step-wise evaluation, each factor was retained if its p-value was <0.10 or if its inclusion modified the effect of HHV-6 by >10%. Potential interactions between HHV-6 and other key variables were formally assessed when indicated. Interaction terms between HHV-6 and aGVHD and were evaluated in mortality models. We were unable to explore interaction between HHV-6 and stem cell source because most cord blood transplant patients had HHV-6 reactivation. Instead, analyses were repeated on a cohort without the cord blood transplant recipients. All reported p-values are two sided, and considered significant if p<0.05.
Results
Of the 315 patients included in analyses, 63 (20%) were <21 years of age, 169 (54%) received myeloablative conditioning, and 217 (69%) received cells from unrelated or HLA-mismatched related donors (Table 1). The stem cell source was growth factor-mobilized peripheral blood cells for 233 (74%), bone marrow for 61 (19%), and cord blood for 21 (7%). Of the 21 cord blood transplant recipients, 76% received double cord blood transplants.
HHV-6
HHV-6 was detected in 111 (35%) of the 315 patients by day 84 post-HCT. The median time to first detection among those with reactivation was 20 [intraquartile range (IQR) 15, 28] days after HCT. The median duration of the first episode of HHV-6 detection, from the first positive through the last consecutive positive, was 4 days (IQR 1, 11). The median maximum DNA level was 873 copies/mL (IQR 175, 4580), and the HHV-6 DNA level was ≥ 1,000 copies/mL in 59 (17%). HHV-6 was detected in 16 (76%) of the 21 cord blood transplant recipients, and in all cases, the HHV-6 DNA level was ≥ 1,000 copies/mL. All detected HHV-6 was type B.
HHV-6 and CMV
Only the patients with donor or recipient CMV seropositive status (n=203) were included in analyses of CMV reactivation. The distribution of HLA match categories and stem cell source among these 203 patients was similar to the overall cohort (data not shown). Any level of CMV reactivation occurred in 128 (63%) of 203 patients who were donor or recipient CMV-seropositive. High-level CMV reactivation (antigenemia ≥ 10 cells per high-powered field or CMV DNA level ≥1000 copies /mL) occurred in 54 (26%), and CMV disease occurred in 7 (3%). First detection of any level of CMV reactivation occurred at a median of 36 (IQR 25, 53) days after HCT. In donor or recipient CMV seropositive patients, HHV-6 reactivation (any DNA level) was independently associated with increased risk of subsequent CMV reactivation of any level (aHR 1.88, 95% CI 1.26-2.82, p = 0.002), but was not associated with high-level CMV reactivation (Table 2). In contrast, high-level HHV-6 reactivation (>1,000 copies DNA/mL) was not significantly associated with CMV reactivation at any level (Table 2), but it was strongly associated with increased risk of subsequent high-level CMV reactivation (aHR 3.11, 95% CI 1.52-6.36, p = 0.002). The addition of grades II-IV or grades III-IV aGVHD to the multivariable models of HHV-6 and CMV reactivation did not markedly change the HR point estimates or statistical significance for HHV-6 even though the aGVHD covariates were strong predictors of CMV reactivation (data not shown). Analyses were repeated excluding the cord blood transplant recipients, and the results were not meaningfully different (data not shown). The number of observations was insufficient to evaluate HHV-6 reactivation as a predictor of CMV disease. In addition, there were few patients who received ganciclovir, foscarnet, or cidofovir early after HCT; therefore we were unable to evaluate the impact of these antivirals on risk of CMV or HHV-6 reactivation or HHV-6 viral DNA levels.
Table 2
Multivariable models evaluating HHV-6 as a predictor of CMV reactivation by day 100 following HCT in patients seropositive for CMV (donor or recipient).
| Any level CMV | High-level CMV | |||
|---|---|---|---|---|
| aHR (95% CI) | P | aHR (95% CI) | P | |
| Models with any level HHV-6 | ||||
| HHV-6 reactivation | 1.88 (1.26- 2.82) Covariatesabce | 0.002 | 1.14 (0.61-2.13) Covariatesb, c, e | 0.69 |
| Models with high level HHV-6 | ||||
| HHV-6 reactivation | 1.56 (0.96- 2.54) Covariatesabce | 0.08 | 3.11 (1.52- 6.36) Covariatesb,c,e,f | 0.002 |
Open in a separate window
aHR = adjusted hazards ratio, CI = confidence interval
Covariates included in the final multivariable models:
arecipient CMV positive
bstem cell source
cmyeloablative transplant dpre-transplant lymphocyte count
eHLA match
fage
HHV-6 and aGVHD
Grades II-IV aGVHD was diagnosed in 240 (76%) patients with onset of symptoms or signs at a median of 27 (IQR 19, 41) days after transplantation. Grades III-IV aGVHD was diagnosed in 50 (16%) patients with onset of symptoms at a median of 22 (IQR 12, 28) days after transplantation. Among the subjects who developed HHV-6 and subsequent aGVHD, the interval between the first detection of HHV-6 and onset of the first symptoms or signs of grades II-IV aGVHD was a median of 11 (IQR 5, 25) days (n = 63) and between HHV-6 and grades III-IV aGVHD was a median of 4.5 (IQR 3, 7) days (n = 12). Similar intervals were observed between high-level HHV-6 and the aGVHD outcomes (data not shown). In a multivariable model, we observed a non-significant but suggestive association between HHV-6 reactivation subsequent grades II-IV overall aGVHD (aHR 1.36, 95% CI 0.99-1.88, p = 0.06). The association was strengthened when high-level HHV-6 reactivation was evaluated (aHR 2.39, 95% CI 1.60-3.56, p < 0.001). HHV-6 reactivation was not, however, significantly associated with grades III-IV overall aGVHD (Figure 2).
Figure 2
Results from multivariable models evaluating HHV-6 reactivation as a risk factor for subsequent aGVHD by day 100. A. Full cohort. B. Excluding Cord Blood recipients. Covariates included in the final multivariable models: 1age, 2conditioning regimen, 3stem cell source, 4sex, 5underlying disease, 6HLA match, 7CMV serostatus, 8CD34 dose, 9co-morbidity index, 10female donor/male recipient.
HHV-6 reactivation was evaluated in multivariable models as a risk factor for organ-specific sub-types of aGVHD: skin, gastrointestinal, and liver aGVHD. A significant association was demonstrated between HHV-6 and stages 3-4 skin aGVHD (aHR 1.71, 95% CI 1.04-2.81, p = 0.04), and the association was strengthened when high-level HHV-6 reactivation was evaluated (aHR 2.01, 95% CI 1.08-3.75, p = 0.03) (Figure 2). High-level HHV-6 was associated with increased risk of both subsequent stages 1-4 and 2-4 hepatic aGVHD (aHR 2.34, 95% CI 1.03-5.30, p = 0.04 and aHR 4.53, 95% CI 1.10-18.68, p = 0.04, respectively) (Figure 2). High-level HHV-6 was also associated with increased risk of subsequent stages 1-4 gastrointestinal aGVHD (aHR 1.68, 95% CI 1.05- 2.68, p = 0.03) (Figure 2)
Analyses of aGVHD excluding cord blood transplant recipients produced similar results (Figure 2), although associations became statistically significant for any level of HHV-6 and grades II-IV overall aGVHD as well as for high-level HHV-6 and grades III-IV overall aGVHD. In addition, statistical significance was diminished for associations of high-level HHV-6 and stages 1-4 hepatic aGVHD as well as for both any level and high-level HHV-6 and stages 3-4 skin aGVHD.
HHV-6 and cGVHD
HHV-6 reactivation was not associated with cGVHD by day 100 in either univariate (HR 1.19, 95% CI 0.84-1.69, p = 0.33) or multivariate (HR 1.18, 95% CI 0.82-1.71, p = 0.38) analyses. Similar results were obtained at the 1-year time point (data not shown).
HHV-6 and Mortality
Overall, 60 (19%) of the 315 patients died by day 200 after HCT and in 40 (13%), cause of death was considered to be non-relapse mortality. Of the 40 patients with non-relapse mortality, 17 (43%) had a fatal infection with (n=7) or without (n=10) GVHD. Other causes included: multiorgan failure (10; 25%), respiratory failure (3; 8%), cerebrovascular event (3; 8%), and other causes (7; 18%).
High-level HHV-6 reactivation was independently associated with increased risk of non-relapse mortality (aHR 2.70, 95% CI 1.15-6.34, p=0.02). There was a suggestion that both any level of HHV-6 reactivation and high-level HHV-6 reactivation were associated with an increased risk of overall mortality although these findings were not statistically significant (Table 3). Significant interactions were observed between high-level HHV-6 and grades III-IV overall aGVHD for both overall survival (p=0.04) and non-relapse mortality (p=0.05). Thus, we created a composite variable for high-level HHV-6 and grades III-IV aGVHD for use in the mortality models to illustrate the varying degrees of association depending on whether each factor was present singly or in combination (Table 3). When they occurred alone, both high-level HHV-6 and grades III-IV aGVHD were associated with increased risks of mortality compared to absence of the two conditions. Co-occurrence of high-level HHV-6 and grades III-IV aGVHD was also associated with increased risk of mortality compared to subjects with neither condition; however, the level of risk was not significantly different from the risk associated with each condition alone, that is among patients with grades III-IV aGVHD, addition of high-level HHV-6 did not result in a significantly increased risk of mortality, although the number of subjects who experienced both high-level HHV-6 and grades III-IV aGVHD was relatively small (n = 13). Inclusion of grade II-IV aGVHD in the models did not have a large effect on the associations between HHV-6 and mortality. CMV reactivation was not associated with risk of subsequent mortality and did not significantly alter the associations between HHV-6 reactivation and mortality (data not shown). Analyses of mortality were also performed excluding the cord blood transplant recipients. Point estimates for the HRs were similar, but significance was diminished or lost across most categories (data not shown). Only high level HHV-6 remained significantly associated with non-relapse mortality (HHV-6 only: HR 3.13, 95% CI 1.22-8.03, p 0.02).
Table 3
Multivariable models evaluating HHV-6 (modeled as any level and level >1000 copies/mL) as a risk factor the outcome mortality (evaluated as overall mortality and as non-relapse mortality). Models were evaluated with and without aGVHD as a covariate (modeled as grade 2-4 or grade 3-4) and with and without CMV reactivation as a covariate. CMV reactivation was not a significant predictor of mortality and its inclusion in the model did not effect the HHV-6 point estimate or significance (data not shown).
| Overall mortality by day 200 | Non-relapse mortality by day 200 | |||
|---|---|---|---|---|
| aHR (95% CI) | P | aHR (95% CI) | P | |
| Models with any level of HHV-6 | ||||
| Any HHV-6 alone | ||||
| • Any HHV-6 | 1.69 (0.99-2.87) Covariates abc | 0.053 | 1.74 (0.88-3.41) Covariates abcdefg | 0.11 |
| Any HHV-6 with aGVHD grades II-IV | ||||
| • Any HHV-6 | 1.60 (0.93-2.75) | 0.09 | 1.73 (0.87-3.42) | 0.12 |
| • aGVHD grades II-IV | 1.36 (0.71-2.58) Covariates abc | 0.35 | 1.61 (0.72-3.62) Covariates a, c, f | 0.25 |
| Any HHV-6 with aGVHD grades III-IV | ||||
| • Any HHV-6 | 1.42 (0.82-2.45) | 0.21 | 1.37 (0.70-2.70) | 0.36 |
| • aGVHD grades III-IV | 2.88 (1.61-5.14) Covariates abc | <0.001 | 4.39 (2.23-8.63) Covariates a, c | <0.001 |
| Models with high-level HHV-6 | ||||
| HHV-6 alone | ||||
| • HHV-6>1000c/mL | 1.95 (0.99-3.86) Covariates abcdef | 0.054 | 2.70 (1.15-6.34) Covariates a, cdef, h | 0.02 |
| HHV-6 with aGVHD grades II-IV | ||||
| • HHV-6>1000c/mL | 1.81 (0.90-3.62) | 0.10 | 2.50 (1.04-6.03) | 0.04 |
| • aGVHD grades II-IV | 1.33 (0.69-2.57) Covariates abcd,f | 0.39 | 1.35 (0.59-3.09) Covariates a, cdef, h | 0.48 |
| HHV-6 with aGVHD grades III-IV i | ||||
| • HHV-6 & aGVHD grades III-IV composite variable | ||||
| ○ Neither HHV-6 or aGVHD grades III-IV | Reference | Reference | ||
| ○ HHV-6>1000c/mL only | 2.61 (1.22-5.62) | 0.014 | 3.61 (1.45-8.99) | 0.006 |
| ○ aGVHD grades III-IV only | 4.42 (2.34-8.34) | <0.001 | 7.02 (3.35-14.7) | <0.001 |
| ○ HHV-6>1000c/mL & aGVHD grades III-IV | 2.50 (0.72-8.71) Covariates abcd, f | 0.15 | 4.36 (0.93-20.4) Covariates a,c,f,h | 0.06 |
Open in a separate window
aHR = adjusted hazard ratio, CI = confidence interval
Other covariates included in the final models:
aage
bunderlying disease
cCMV serostatus
dmyeloablative conditioning
eco-morbidity index
fstem cell source
gchronic gvhd
hrace 28
iDue to the significant interaction between high-level HHV-6 reactivation and aGVHD grades III-IV for each mode (p=0.04 for Overall Survival and p=0.05 for NRM)l, a 4-level composite variable was evaluated in the final models including these factors to illustrate the varying association depending on whether each factor is present singly or in combination. The reference category for this variable is defined by having neitherHHV-6 or aGVHD grades III-IV. No effect modification was observed between any level HHV-6 and aGVHD grades II-IV, any level HHV-6 and aGVHD grades III-IV, or between high level HHV-6 and aGVHD gr II-IV. Therefore, a composite variable was not required for these models.
HHV-6 and the composite endpoint
HHV-6 reactivation of any level was associated with the composite endpoint of any mortality, grades II-IV aGVHD or any level of CMV reactivation (aHR 1.47, 95% CI 1.08-2.01, p = 0.015). Other variables retained in the model included gender, pre-HCT CMV seropositivity in the donor or recipient, type of conditioning regimen, stem cell source, pre-transplant lymphopenia and HLA match between the donor and recipient. High-level HHV-6 appeared even more strongly associated with this composite outcome (aHR 2.02, 95%CI 1.35-3.01, p<0.001). Patient age, stem cell source, type of conditioning regimen, HLA match between the donor and recipient, and pre-HCT positive serology for HSV remained in the final model.
Discussion
In this prospective study of 315 HCT recipients, we found that HHV-6 reactivation was associated with increased risk of subsequent of CMV reactivation, aGVHD, and mortality. Associations between HHV-6 and grades 2-4 aGVHD, non-relapse mortality, and high-level CMV reactivation were strengthened when high-level HHV-6 was evaluated.
Previous studies have demonstrated an association between HHV-6 and aGVHD[2,17] of which several have used multivariable analyses in an attempt to control for potential confounders.[4,6,7] In the current study, we also evaluated organ-specific subtypes of aGVHD as outcomes. Results from these analyses suggest that HHV-6 may play a role in development of each of the aGVHD subtypes. It is interesting that the strength of the association between high-level HHV-6 and skin aGVHD was lessened when cord blood transplant recipients were removed from the analysis. Most (76%) of the 21 cord blood transplant recipients in our study received double cord blood transplants, and these transplants have been associated with greater risk of skin aGVHD.[18] This finding, along with the fact that 76% of the cord blood transplant recipients in our study had high level HHV-6 reactivation would explain these results. The association of HHV-6 reactivation with skin aGVHD raises the question of whether HHV-6 might actually trigger aGVHD or simply cause symptoms and signs that are then attributed to aGVHD. Others have demonstrated an association between HHV-6 and rash during the first month after HCT[19-22] and HHV-6 is known to cause rash during primary acquisition.[23] Confirming the role of HHV-6 in aGVHD and distinguishing rash related to HHV-6 from rash related to aGVHD is necessary in order to target therapy appropriately.
While a causal association between HHV-6 and aGVHD has not been established, certain data support the biologic plausibility of such a relationship. Specifically, in vitro and limited clinical data suggest that HHV-6 infection or reactivation may cause a pro-inflammatory or type I immune response, which may play an important role in the development of aGVHD. For example, a type I immune response polarization has been observed during HHV-6 infection of T cells in vitro,[24] and small studies of HCT recipients have documented a pro-inflammatory cytokine response (primarily elevated IL-6 concentrations) associated with HHV-6 reactivation.[25,26] In parallel, a number of pro-inflammatory or type I cytokines (IL-2,sIL-2R, IL-5, IL-6, IL-10, IFN-gamma, IL-12, and IL-18) have been associated with aGVHD.[27-30] Although few studies exploring potential associations between HHV-6, the immune response, and aGVHD have been described, data from a small number of patients with HHV-6 reactivation after HCT have implied a temporal association between elevated cytokines and rash or aGVHD.[26] Further in vivo study of the effects of HHV-6 on the immune system may help elucidate the pathogenesis of HHV-6 related disease.
Seemingly in contrast to the concept of an HHV-6-induced pro-inflammatory response, other in vitro evidence suggests that HHV-6 reactivation might suppress the antiviral immune response, potentially through suppression of IL-12 production.[31-33] Such suppression of antiviral immune responses could potentially promote the HHV-6 infection itself, as well as other viral infections, including CMV. In solid organ transplant recipients, HHV-6 reactivation has been independently associated with CMV reactivation and disease;[34,35] while studies conducted in HCT populations have yielded variable results. In a study of 21 allogeneic HCT recipients, HHV-6 reactivation was associated with an absence of CMV-specific lymphocyte proliferative responses, and persistence of HHV-6 detection was associated with need for repeated courses of preemptive antiviral therapy against CMV during the first 6 months after HCT.[36] In contrast, a study of 68 allogeneic HCT recipients found that HHV-6 reactivation was associated with CMV reactivation in univariate analysis but not multivariable analyses.[5] In addition, detection of HHV-6 DNA in plasma did not appear to affect CMV-specific T-cell immunity reconstitution as measured by intracellular cytokine staining. Therefore, the authors concluded that the severe immunosuppression that attends HCT leads to HHV-6 and CMV reactivation but that HHV-6 does not predispose to CMV reactivation or influence the course of active CMV infection.[5] There have been few published studies that have investigated the association between HHV-6 reactivation and CMV reactivation/ disease in HCT recipients. The inconsistent results may in part be due to the small size of previous studies. Given the hazard ratio that we observed for the association between HHV-6 and CMV reactivation, relatively large studies, involving several hundred subjects such as in our study, are needed to address this question. Despite having a large study, we did not have an adequate number of patients with HHV-6 and subsequent receipt of ganciclovir or foscarnet to assess the impact of these antivirals on HHV-6 levels and/or on CMV reactivation. A trial testing the effect of an HHV-6-active antiviral on HHV-6 and CMV reactivation would likely clarify these issues.
In the present study we also found that HHV-6 reactivation was independently associated with non-relapse mortality, and borderline associations were noted between any- and high-level HHV-6 reactivation and all-cause mortality. The association between HHV-6 reactivation and mortality in HCT has been reported previously in only a few studies.[4,6] Our data suggests that that the association between HHV-6 and mortality may be mediated partly through aGVHD. Many complex pathways could link HHV-6 reactivation with increased risk of CMV reactivation, aGVHD, and mortality. It is also possible that other clinical conditions, such as central nervous system disease, pneumonitis, or bone marrow suppression might mediate the relationship between HHV-6 and mortality. The experimental design offered by a randomized controlled trial would likely shed light on these relationships. Our results also suggest that cord blood transplant recipients were important in driving the statistical significance of the associations we observed between HHV-6 and mortality. Thus, cord blood recipients may be an important group to target for future intervention studies.
This study has many strengths, including the size of the cohort, the frequency and regularity of HHV-6 testing, and the systematic collection of outcome data. This allowed us to address important questions regarding the possible association of HHV-6 reactivation with CMV reactivation, GVHD, or mortality more definitively than previous smaller studies. However, due to the observational study design, we cannot conclude that HHV-6 reactivation is causally related to these clinical outcomes. The ubiquitous and persistent nature of HHV-6 infection poses significant challenges in establishing causal associations between viral reactivation and disease, especially in immunocompromised populations with multiple complex medical problems. An antiviral intervention trial would provide the experimental data required to determine causality.
In summary, we demonstrated an independent and quantitative association between HHV-6 reactivation and the outcomes of CMV reactivation, aGVHD, and mortality. A randomized antiviral trial is warranted to determine if reducing HHV-6 reactivation will reduce the incidence of these outcomes after HCT.
Acknowledgements
We are grateful to the patients and their families who participated in this study; without their patience and dedication, this study could not have been undertaken. In addition, we would like to thank Mohamed Sorror, MD for sharing data regarding the HCT co-morbidity index, Brenda Sandmaier, MD for sharing data regarding progression which was used for cause of death in non-myeloablative patients, and Anna Rashevsky for her excellent laboratory work.
Funding
This work was supported by National Institutes of Health [grant numbers AI057639, CA018029, HL036444, HL94294, and CA078902.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Financial Disclosure
MB reports receiving research funding for clinical trials and consulting fees from Roche/Genentech and Chimerix Inc. All other others have no disclosures.
References
1. Okuno T, Takahashi K, Balachandra K, et al. Seroepidemiology of human herpesvirus 6 infection in normal children and adults. J Clin Microbiol. 1989;27:651–3. [PMC free article] [PubMed] [Google Scholar]
2. Hentrich M, Oruzio D, Jager G, et al. Impact of human herpesvirus-6 after haematopoietic stem cell transplantation. Br J Haematol. 2005;128:66–72. [PubMed] [Google Scholar]
3. Imbert-Marcille BM, Tang XW, Lepelletier D, et al. Human herpesvirus 6 infection after autologous or allogeneic stem cell transplantation: a single-center prospective longitudinal study of 92 patients. Clin Infect Dis. 2000;31:881–6. [PubMed] [Google Scholar]
4. Zerr DM, Corey L, Kim HW, Huang ML, Nguy L, Boeckh M. Clinical outcomes of human herpesvirus 6 reactivation after hematopoietic stem cell transplantation. Clin Infect Dis. 2005;40:932–40. [PubMed] [Google Scholar]
5. Tormo N, Solano C, de la Camara R, et al. An assessment of the effect of human herpesvirus-6 replication on active cytomegalovirus infection after allogeneic stem cell transplantation. Biol Blood Marrow Transplant. 2010;16:653–61. [PubMed] [Google Scholar]
6. de Pagter PJ, Schuurman R, Visscher H, et al. Human herpes virus 6 plasma DNA positivity after hematopoietic stem cell transplantation in children: an important risk factor for clinical outcome. Biol Blood Marrow Transplant. 2008;14:831–9. [PubMed] [Google Scholar]
7. Wang LR, Dong LJ, Zhang MJ, Lu DP. Correlations of human herpesvirus 6B and CMV infection with acute GVHD in recipients of allogeneic haematopoietic stem cell transplantation. Bone Marrow Transplant. 2008;42:673–7. [PubMed] [Google Scholar]
8. Zerr DM, Fann JR, Breiger D, et al. HHV-6 reactivation and its effect on delirium and cognitive functioning in hematopoietic cell transplantation recipients. Blood. 2011;117:5243–9. [PMC free article] [PubMed] [Google Scholar]
9. Clark DA, Nacheva EP, Leong HN, et al. Transmission of integrated human herpesvirus 6 through stem cell transplantation: implications for laboratory diagnosis. J Infect Dis. 2006;193:912–6. [PubMed] [Google Scholar]
10. Boeckh M, Bowden RA, Gooley T, Myerson D, Corey L. Successful modification of a pp65 antigenemia-based early treatment strategy for prevention of cytomegalovirus disease in allogeneic marrow transplant recipients. Blood. 1999;93:1781–2. [PubMed] [Google Scholar]
11. Boeckh M, Ljungman P. How we treat cytomegalovirus in hematopoietic cell transplant recipients. Blood. 2009;113:5711–9. [PMC free article] [PubMed] [Google Scholar]
12. Bensinger WI, Martin PJ, Storer B, et al. Transplantation of bone marrow as compared with peripheral-blood cells from HLA-identical relatives in patients with hematologic cancers. N Engl J Med. 2001;344:175–81. [PubMed] [Google Scholar]
13. Sorror ML, Giralt S, Sandmaier BM, et al. Hematopoietic cell transplantation specific comorbidity index as an outcome predictor for patients with acute myeloid leukemia in first remission: combined FHCRC and MDACC experiences. Blood. 2007;110:4606–13. [PMC free article] [PubMed] [Google Scholar]
14. Leisenring WM, Martin PJ, Petersdorf EW, et al. An acute graft-versus-host disease activity index to predict survival after hematopoietic cell transplantation with myeloablative conditioning regimens. Blood. 2006;108:749–55. [PMC free article] [PubMed] [Google Scholar]
15. Filipovich AH, Weisdorf D, Pavletic S, et al. National Institutes of Health consensus development project on criteria for clinical trials in chronic graft-versus-host disease: I. Diagnosis and staging working group report. Biol Blood Marrow Transplant. 2005;11:945–56. [PubMed] [Google Scholar]
16. Zerr DM, Gupta D, Huang ML, Carter R, Corey L. Effect of antivirals on human herpesvirus 6 replication in hematopoietic stem cell transplant recipients. Clin Infect Dis. 2002;34:309–17. [PubMed] [Google Scholar]
17. Appleton AL, Sviland L, Peiris JS, et al. Human herpes virus-6 infection in marrow graft recipients: role in pathogenesis of graft-versus-host disease. Newcastle upon Tyne Bone Marrow Transport Group. Bone Marrow Transplant. 1995;16:777–82. [PubMed] [Google Scholar]
18. MacMillan ML, Weisdorf DJ, Brunstein CG, et al. Acute graft-versus-host disease after unrelated donor umbilical cord blood transplantation: analysis of risk factors. Blood. 2009;113:2410–5. [PMC free article] [PubMed] [Google Scholar]
19. Yoshikawa T, Ihira M, Ohashi M, et al. Correlation between HHV-6 infection and skin rash after allogeneic bone marrow transplantation. Bone Marrow Transplant. 2001;28:77–81. [PubMed] [Google Scholar]
20. Yoshikawa T, Asano Y, Ihira M, et al. Human herpesvirus 6 viremia in bone marrow transplant recipients: clinical features and risk factors. J Infect Dis. 2002;185:847–53. [PubMed] [Google Scholar]
21. Volin L, Lautenschlager I, Juvonen E, Nihtinen A, Anttila VJ, Ruutu T. Human herpesvirus 6 antigenaemia in allogeneic stem cell transplant recipients: impact on clinical course and association with other beta-herpesviruses. Br J Haematol. 2004;126:690–6. [PubMed] [Google Scholar]
22. Wang LR, Dong LJ, Zhang MJ, Lu DP. The impact of human herpesvirus 6B reactivation on early complications following allogeneic hematopoietic stem cell transplantation. Biol Blood Marrow Transplant. 2006;12:1031–7. [PubMed] [Google Scholar]
23. Zerr DM, Meier AS, Selke SS, et al. A population-based study of primary human herpesvirus 6 infection. N Engl J Med. 2005;352:768–76. [PubMed] [Google Scholar]
24. Mayne M, Cheadle C, Soldan SS, et al. Gene expression profile of herpesvirus-infected T cells obtained using immunomicroarrays: induction of proinflammatory mechanisms. J Virol. 2001;75:11641–50. [PMC free article] [PubMed] [Google Scholar]
25. Ogata M, Satou T, Kawano R, et al. Correlations of HHV-6 viral load and plasma IL-6 concentration with HHV-6 encephalitis in allogeneic stem cell transplant recipients. Bone Marrow Transplant. 2010;45:129–36. [PubMed] [Google Scholar]
26. Fujita A, Ihira M, Suzuki R, et al. Elevated serum cytokine levels are associated with human herpesvirus 6 reactivation in hematopoietic stem cell transplantation recipients. J Infect. 2008;57:241–8. [PubMed] [Google Scholar]
27. Ju XP, Xu B, Xiao ZP, et al. Cytokine expression during acute graft-versus-host disease after allogeneic peripheral stem cell transplantation. Bone Marrow Transplant. 2005;35:1179–86. [PubMed] [Google Scholar]
28. Mohty M, Blaise D, Faucher C, et al. Inflammatory cytokines and acute graft-versus-host disease after reduced-intensity conditioning allogeneic stem cell transplantation. Blood. 2005;106:4407–11. [PubMed] [Google Scholar]
29. Shaiegan M, Iravani M, Babaee GR, Ghavamzadeh A. Effect of IL-18 and sIL2R on aGVHD occurrence after hematopoietic stem cell transplantation in some Iranian patients. Transpl Immunol. 2006;15:223–7. [PubMed] [Google Scholar]
30. Fujii N, Hiraki A, Aoe K, et al. Serum cytokine concentrations and acute graft-versus-host disease after allogeneic peripheral blood stem cell transplantation: concurrent measurement of ten cytokines and their respective ratios using cytometric bead array. Int J Mol Med. 2006;17:881–5. [PubMed] [Google Scholar]
31. Smith A, Santoro F, Di Lullo G, Dagna L, Verani A, Lusso P. Selective suppression of IL-12 production by human herpesvirus 6. Blood. 2003;102:2877–84. [PubMed] [Google Scholar]
32. Smith AP, Paolucci C, Di Lullo G, Burastero SE, Santoro F, Lusso P. Viral replication-independent blockade of dendritic cell maturation and interleukin-12 production by human herpesvirus 6. J Virol. 2005;79:2807–13. [PMC free article] [PubMed] [Google Scholar]
33. Lusso P. HHV-6 and the immune system: mechanisms of immunomodulation and viral escape. J Clin Virol. 2006;37(Suppl 1):S4–10. [PubMed] [Google Scholar]
34. Humar A, Malkan G, Moussa G, Greig P, Levy G, Mazzulli T. Human herpesvirus-6 is associated with cytomegalovirus reactivation in liver transplant recipients. J Infect Dis. 2000;181:1450–3. [PubMed] [Google Scholar]
35. Humar A, Kumar D, Caliendo AM, et al. Clinical impact of human herpesvirus 6 infection after liver transplantation. Transplantation. 2002;73:599–604. [PubMed] [Google Scholar]
36. Wang FZ, Larsson K, Linde A, Ljungman P. Human herpesvirus 6 infection and cytomegalovirus-specific lymphoproliferative responses in allogeneic stem cell transplant recipients. Bone Marrow Transplant. 2002;30:521–6. [PubMed] [Google Scholar]
