Prevalence of Chromosomally Integrated Human Herpesvirus 6 (HHV-6) in Patients with HHV-6 Central Nervous System Dysfunction

Abstract

We identified 37 hematopoietic cell transplantation recipients with human herpesvirus 6 (HHV-6) central nervous system dysfunction and tested donors/recipients for chromosomally integrated (ci)HHV-6. One patient had ciHHV-6A with possible HHV-6A reactivation and encephalitis. There was no ciHHV-6 enrichment in this group, but larger studies are needed to determine if patients with ciHHV-6 are at increased risk for HHV-6-associated diseases or other complications.

INTRODUCTION

Human herpesvirus 6 (HHV-6) has a unique ability to integrate into chromosomal telomeres of infected cells¹. When this occurs in germ cells, Mendelian inheritance results in offspring with chromosomally integrated (ci)HHV-6 in every nucleated cell. This condition is present in ~1% of people¹, can be the source of viral reactivation^2,3, and has been implicated in HHV-6 central nervous system (CNS) disease^4,5. To explore whether patients with ciHHV-6 have increased risk for HHV-6-associated disease, we tested the prevalence of ciHHV-6 in allogeneic hematopoietic cell transplantation (HCT) recipients with HHV-6 CNS dysfunction.

We identified 37 allogeneic HCT recipients at our center with HHV-6 CNS dysfunction as previously described⁶. We tested archived samples from donor/recipient pairs for ciHHV-6 using a novel method to detect HHV-6 and human ribonuclease P (RPP30, a reference gene for cell count) DNA by droplet digital PCR (Fig. 1)⁷. CiHHV-6 was ruled out if the ratio of HHV-6 DNA to cell genome equivalents (two RPP30/cell) fell outside the range of 1 ± 0.07.

Figure 1.

Flow diagram for chromosomally integrated HHV-6 testing. Archived pre-HCT patient and donor Beta-lymphoblastoid cell lines were the primary source for ciHHV-6 testing. When unavailable, other appropriate specimens (serum or bone marrow) were used as surrogates.

*Surrogate for donor sample in cases that were missing pre-HCT donor sample. These were obtained post-engraftment.

^†One patient had a HHV-6/cellular DNA ratio significantly >1, making it impossible to rule out ciHHV-6 coupled with reactivation.

Abbreviations: HCT, hematopoietic cell transplantation; HHV-6, human herpesvirus 6; CNS, central nervous system; B-lymphoblastoid, Beta-lymphoblastoid; ciHHV-6, chromosomally integrated HHV-6.

CiHHV-6 species A was detected in one pre-HCT patient sample (prevalence 2.7%; 95% confidence interval, 0.07-14.5%) from a 53 year old man who developed findings consistent with HHV-6 encephalitis after a myeloablative matched, related donor HCT. He initially developed hallucinations and agitation on day 11 after HCT (D+11), one day after starting methylprednisolone 2 mg/kg/day for skin rash. Cerebrospinal fluid (CSF) on D+12 had two lymphocytes, four red blood cells, and protein 43 mg/dl. Bacterial, mycobacterial, and fungal stains and cultures were negative, as was PCR for other herpesviruses. Semi-quantitative PCR for HHV-6 DNA was positive with 2,500 copies/ml. The patient was started on foscarnet 90 mg/kg IV Q12h on D+14 for HHV-6 encephalitis.

The patient had initial improvement in mental status followed by progressive encephalopathy. He engrafted on D+17 and was weaned off steroids by D+23. Repeat CSF testing on D+26 had 2 lymphocytes, 54 red blood cells, and 2,500 copies/ml HHV-6 DNA by semi-quantitative PCR. Additional testing as above was negative. Foscarnet was discontinued on D+29 due to acute kidney injury. Chimerism and flow cytometry studies of bone marrow and blood on D+29 were of donor origin, and PCR for HHV-6 in bone marrow was negative at that time. Brain magnetic resonance imaging was negative on D+33. The patient developed hypoxia on D+38 attributed to aspiration in the setting of obtundation and died on D+40 due to progressive pulmonary disease. Autopsy revealed increased brain microglia and perivascular lymphocyte cuffing, as well as scattered focal hippocampal neuronal dropout; although non-specific, this could be consistent viral encephalitis.

Retrospective testing of multiple archived serum samples with droplet digital PCR detected HHV-6 DNA with decreasing HHV-6/cellular DNA ratios of 0.5 to 0.02 (day 5 through 39). This could be consistent with allograft replacement of recipient ciHHV-6 hematopoietic cells or treatment of HHV-6 reactivation with foscarnet. Sanger sequencing of a 900 base pair region of the HHV-6 envelope glycoprotein B (gB) gene in a pre-HCT cell sample and post-HCT serum sample (D+39) revealed complete homology but divergence from 14 other strains⁸ (Fig. 2). A fresh-frozen brain section of the inferior temporal lobe had HHV-6/cellular DNA ratio of 1.0, consistent with ciHHV-6 but without evidence of additional HHV-6 replication. Testing of formalin-fixed, paraffin-embedded samples from the medial temporal lobes (the predominant site of HHV-6 CNS disease) yielded inconclusive results due to poor preservation. RNA testing and viral culture could not be performed on archived specimens.

Figure 2.

Phylogenetic analysis of HHV-6A gB gene sequences from pre and post-HCT samples of the described patient with ciHHV-6A compared to ciHHV-6A from 4 patients at our center, the Hector-2 cell line (Bioworld Consulting Laboratories), and 9 published sequences⁸.

This figure demonstrates that the sequence of an ~900 base pair region of the HHV-6A gB gene in the discussed patient’s pre-HCT Beta-lymphoblastoid cell line and post-HCT (D+39) serum sample were identical and diverged from 14 other available sequences.

Bar = number of nucleotide changes per 100 sites.

The gB gene sequences of 9 HHV-6A primary or laboratory-adapted isolates were obtained from Dr Henri Agut (Paris, France), and the geographical origins were as follows: GS, United States; U1102, Uganda; SIE, Côte d’Ivoire; TAN, Congo; and 1120, E540_132, 1116, 719 and, p523, France. An additional 4 isolates were obtained from patients at the University of Washington with chromosomally integrated HHV-6: UWciHHV6_1, UWciHHV6_2, UWciHHV6_3, UWciHHV6_4.

The tree was constructed from the comparison of gB gene nucleotide sequences using the free Phylogeny.fr webservice, available at http://www.phylogeny.fr/version2_cgi/index.cgi.

This is the first epidemiologic study of the prevalence of ciHHV-6 in patients with HHV-6-associated CNS dysfunction. Convincing evidence of HHV-6 reactivation and pathogenicity from ciHHV-6 cell lines is well-described^2-5. We did not identify a clear overrepresentation of ciHHV-6 in this cohort of 37 patients compared to population-based studies¹, but the confidence interval was wide and suggests an incidence as high as 14%. We also describe the first case of a patient with ciHHV-6A and HHV-6A reactivation as a possible cause of post-HCT encephalitis.

Although HHV-6A DNA detected in post-HCT samples may have represented cells with ciHHV-6A rather than active replication, and alternative causes of encephalitis were possible, we think the findings are best explained by HHV-6A reactivation. De novo HHV-6 infection was unlikely given sequence homology between pre and post-HCT HHV-6 strains. Recipient hematopoietic cells with ciHHV-6A would be an unlikely source of persistent HHV-6A DNA in serum samples after a myeloablative HCT, especially given evidence of full donor chimerism and absence of HHV-6 DNA detection in the D+29 bone marrow biopsy. HHV-6A DNA in CSF samples was also unlikely to be from CNS or residual hematopoietic cells with ciHHV-6A given the low ratio of nucleated cells (2) to HHV-6 copies (2,500); a study of HHV-6 levels in CSF from patients with ciHHV-6 and unrelated CNS disease showed close correlation between leukocyte count and viral load⁹. Unfortunately, RNA testing and viral culture of saved specimens was not possible due to prior processing and extended storage.

In conclusion, we did not find evidence of ciHHV-6 enrichment in a cohort of HCT patients with HHV-6 CNS dysfunction compared to the general population, but larger studies are needed to comprehensively determine the true incidence and prevalence of reactivation in high-risk patients. Our findings begin to address the call to reevaluate transplant practices in the setting of ciHHV-6¹⁰ and underscore the need for large, multicenter collaborations to determine the impact of ciHHV-6 on patient outcomes.