Human herpes virus-6 (HHV-6) is a β-herpes virus that was first described in children in 1988 as the causative agent of exanthem subitum (1). HHV-6 infects nearly all children by 3 years of age (2) and is now classified into two closely related but distinct species: HHV-6A and HHV-6B (3). This recognition as separate viruses was made by the International Committee on Taxonomy of Viruses in 2012, primarily due to their different epidemiological, biological, and immunological distinctions. Primary HHV-6B infection may be asymptomatic and present with nonspecific symptoms such as fever, fussiness, rash, diarrhea, and seizures in children (1). The epidemiology of HHV-6A, on the other hand, is still largely uncertain, although it is thought to be acquired later in life and may be more commonly associated with severe neurological disease compared with HHV-6B (4). In adults, the virus has been associated with a broad range of clinical syndromes, ranging from mild undifferentiated rashes, temporal lobe epilepsy (5) to multiple sclerosis and encephalitis (6, 7). Unique to HHV-6 is its ability to integrate into chromosomes (8, 9). As a consequence of the presence of the viral genome in every nucleated human cell, patients with chromosomally integrated HHV-6 (ciHHV-6) are characterized by very high levels of HHV-6 DNA in the blood and tissues (10). ciHHV-6 has been misinterpreted as “active” infection, leading to unnecessary antiviral treatment.

The pathogenicity of HHV-6A, HHV-6B, and ciHHV-6 in immunocompromised hosts, such as recipients of hematopoietic cell transplantation (HCT), solid organ transplantation (SOT), and chimeric antigen receptor T-cell (CAR-T) therapy, has been reported. It is assumed that HHV-6B is the more prevalent species, while HHV-6A is suspected as the etiology of most cases of HHV-6 encephalitis with some outliers (11). However, there are also sporadic reports of other clinical syndromes including myelitis, myocarditis, hepatitis, and hemophagocytic syndromes. ciHHV-6, on the other hand, has been associated with an increased risk of acute graft-versus-host disease (aGvHD) in allogeneic hematopoietic transplants, and it has been most often mistaken for active infection (10).

Earlier reviews have focused only on specific hosts (12), virus types (13), or an aspect of HHV-6 (14). Hence, we aimed to perform a more encompassing systematic review to comprehensively examine the epidemiology of HHV-6 in all these populations by describing its different syndromes and patient outcomes.

With the help of a professional librarian, we searched multiple databases (i.e., PubMed, Embase, Scopus, and Medline/Ovid) and identified all cases of HHV-6 after SOT, HCT, and CAR-T from inception to May 31, 2024. Search terms included the keywords “human herpes virus 6 or HHV-6” and “solid organ or hematopoietic transplant” or “encephalitis.” The complete search strategy is shown in Appendix 1. The article references were reviewed for additional cases. Our search was limited to publications in English or those with an English translation. All reports of adult SOT, HCT, and CAR-T recipients who developed HHV-6 after transplantation within the study period were eligible for study inclusion and were screened by two authors (CLA and RRR). Only those cases where the symptoms were attributed to HHV-6 were included. HHV-6 infection was defined as evidence of viremia or DNAemia, while disease was confirmed in the presence of a compatible clinical syndrome [e.g., fever and rash, post-transplant acute limbic encephalitis (PALE), myelitis, myocarditis, and hepatitis, among others] and/or isolation of HHV-6 from culture, shell vial assay, or via molecular methods (e.g., polymerase chain reaction or metagenomic sequencing). The abstracts of case reports were included if sufficient clinical information was provided. Recipients with ciHHV-6 were included only if there was clinical disease. All case reports or series included symptomatic HHV-6, but incidences of asymptomatic (e.g., viremic) or symptomatic HHV-6 as reported by cohort studies were captured and recorded. We excluded pediatric cases and cases diagnosed prior to transplantation or upon removal of the transplant allograft. Reports on HHV-6 where clinical information was inadequate or lacking, could not be extracted in detail, or had mixed populations (e.g., predominantly pediatric or non-transplant) were excluded. Data were extracted (CLA) and coded into an Excel spreadsheet. Informed consent was not required as these cases have already been previously reported. The PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses) was used to help generate the four-phase flow diagram.

Our search identified a total of 1,239 studies across multiple databases (PubMed, n = 122; Embase, n = 987; Medline, n = 85; and Scopus, n = 45). After the reference review (n = 25) and the exclusion of duplicates (n = 24), a total of 164 studies were included in the final review ( Figure 1 ). There were 136 (32 SOT, 100 HCT, 3 CAR-T, and 1 both SOT/HCT) case reports or series (11, 15–149) that contributed a total of 268 unique cases: 225 HCT, 37 SOT, and 6 CAR-T recipients.

A total of 39 cohort studies (34, 42, 45, 51, 58, 94, 126, 131, 138, 143, 146, 147, 149–175) on HCT (28), SOT (9), CAR-T (1), and mixed SOT/HCT (1) recipients were also included. Of these, 13 studies provided detailed clinical cases that were included in our systematic review of case reports or series (34, 42, 45, 51, 58, 94, 126, 131, 138, 143, 146, 147, 149).

Through this systematic review of HHV-6 in SOT, HCT, and CAR-T recipients, we synthesized information from more than 200 cases and 38 cohort studies and draw attention to the following observations: 1) the rates of clinical and asymptomatic reactivation are highly variable; 2) encephalitis is the most common but not the only clinical syndrome; 3) fever occurrence is highly variable, ranging from 30% to 80%; 4) diagnosis is most often secured using molecular methods in addition to compatible clinical disease; 5) MRI is the imaging modality of choice for neurological disease; 6) antiviral drug preference differs between HCT and SOT; and 7) there is a high rate of long-term neurological sequelae.

Based on our review, the incidence of HHV-6 reactivation is variable and can be as high as 95% in either the SOT or the HCT populations. This is not surprising as HHV-6, similarly to other human herpes viruses, exhibits lifelong latency and tends to reactivate with immunosuppression. The high rate of HHV-6 reactivation is documented in studies from centers that have systematically performed HHV-6 PCR testing, either prospectively or retrospectively in stored clinical samples (146, 174, 175). The vast majority of HHV-6 reactivation that was detected through surveillance is asymptomatic and transient. Hence, the monitoring of HHV-6 is controversial, as asymptomatic or subclinical reactivation happens very often, occurring only transiently, with no impact on the clinical outcomes, and leading to unnecessary treatment with antivirals. As shown in Tables 3 and 4 , clinical HHV-6 disease did not occur (138) or occurred in a much smaller proportion of patients, despite the high overall incidence of HHV-6 reactivation in both HCT (94, 140, 146, 151) and SOT recipients (155, 160, 171, 172). Thus, the clinical utility of routine HHV-6 surveillance is debated. More research is needed to identify patients at high risk of clinical disease and to define the optimal timing and frequency of monitoring for HHV-6 infections in these high-risk patients.

This review confirms that the majority of clinical HHV-6 disease (204/268, 76.1%) in transplant and CAR-T populations manifests as a neurological syndrome, primarily encephalitis. These patients presented within the first month of transplantation with varied neurological symptoms, including confusion, amnesia, and seizures, and the majority had abnormal findings on initial imaging. We propose that HHV-6 viremia in the presence of confusion, amnesia, and seizures plus limbic involvement seen on MRI in a post-transplant or CAR-T recipient is defined as probable HHV-6, until proven otherwise. Notably, many of those who developed a neurological syndrome were HHV-6B. While this could be due to reporting bias and the overall higher prevalence of HHV-6B, it highlights that it has a similar neurotrophic property to that of HHV-6A (176). Indeed, in one study, it was proposed that both viruses are equally neurotrophic but that HHV-6A may potentially be more virulent (177). Whether these in vitro data translate into worse outcomes for those infected with HHV-6A remains to be seen. Interestingly, only a small subset of patients (15 HCT and 1 SOT) was specifically reported as HHV-6 PALE. HHV-6 PALE was first described in HCT recipients (11) and has not been formally defined but is the primary involvement of the limbic system driven by HHV-6 infection. One study (148) proposed characterizing it as a distinct entity with acute alterations in mental status, prominent amnesia, and unexplained seizures. Based on this definition, and compatible with the MRI findings, at least 14 other cases fit the criteria for HHV-6 PALE; however, these cases were simply described as “encephalitis.” This is worth mentioning as it highlights the under-recognition of this entity, as well as the need to more clearly define this category of patients. Early recognition of this specific syndrome may have clinical implications for treating clinicians—such as the association with HHV-6, the prevention of progressive disease, and the potential avoidance of neurological sequelae with early antiviral therapy. Among the HCT recipients, the majority with encephalitis were either allogeneic or cord blood recipients, which may be due to their greater level of immunosuppression compared with autologous transplants. The risk factors for HHV-6 reactivation and disease included the use of a mismatched donor and steroids (146, 174). For the patients in this review, approximately half were mismatches or unrelated donors (50/92, 54.3%) or were given steroids for GvHD (76/183, 41.5%). There may be other host factors related to the complex immunological CD8 T-cell responses, which are dampened by transplant-related immunosuppression that may have significantly influenced the HHV-6 reactivation (178); however, these were not explored in this review.

When there is suspicion of encephalitis, we advocate MRI as the preferred imaging of choice. MRI was able to capture findings not initially seen on the CT scan, even when done on the same day. Imaging evidence involving one or both hippocampi and adjacent medial temporal lobe structures of the limbic system, including the amygdalae and parahippocampal gyri, is suggestive of “limbic encephalitis” and should help narrow the differential diagnosis (148). Although other causes such as paraneoplastic malignancies, neurosyphilis, herpes simplex, and varicella zoster have been implicated in limbic encephalitis, HHV-6 should be a leading impression given the right clinical context.

Other syndromes of HHV-6 occurred but were much less common. Disseminated disease was infrequently seen among both SOT and HCT recipients. For HCT recipients, the median time to presentation of disseminated disease was beyond the period of engraftment (e.g., 171 days). Of the reported cases, 5/8 (62.5%) developed aGvHD and were treated with steroids. We propose that this increased level of immunosuppression contributed to the risk of disseminated disease. However, given the limited number, more definite conclusions cannot be made. In contrast, disseminated disease among SOT recipients occurred within the first month of transplantation. Although it is unclear what drove early disseminated disease in these patients, we hypothesize that the choice of induction immunosuppression played a role.

Fever was more frequent among CAR-T recipients and occurred in a third of other patients. As such, for individuals without a fever, other concomitant symptoms of HHV-6-related disease should be sought. Rashes, diarrhea, and cytopenias were commonly reported with other HHV-6 syndromes. Rashes from a viral exanthem such as HHV-6 are difficult to distinguish from aGvHD based on clinical appearance. In most cases, a biopsy is necessary to determine the etiology of the rashes as the approach to treatment differs.

The diagnosis of HHV-6 was secured primarily using molecular methods (e.g., DNA PCR) in the majority of patients. Confirmation via tissue diagnosis was rarely done, even among SOT recipients. Although the isolation of HHV-6 in cell cultures is the reference method and unambiguously demonstrates the presence of infectious viral particles, it is poorly sensitive, time-consuming, expensive, and unavailable in most centers and thus cannot be used for routine diagnosis (179). However, HHV-6 DNAemia by itself can be exceptionally tricky to interpret in the context of ciHHV-6. ciHHV-6 diagnosis previously required hair follicle or molecular cytogenetic analysis, which is impractical in the clinical setting. Experts propose that ciHHV-6 can be assumed in the presence of >1 million genomic copies of HHV-6 per milliliter of whole blood (10). In our review, the 10 ciHHV-6 cases were treated because all had symptoms consistent with clinical disease. However, the literature is replete with reports of asymptomatic high-level HHV-6 viremia (10, 180–182), with delayed realization of the possibility of ciHHV-6 and the need for treatment (10). As such, careful consideration for ciHHV-6 should be made when there is high-level viremia in an asymptomatic patient.

Interestingly, ganciclovir was the primary therapy used in 54.5% (18/33) of SOT recipients, while foscarnet (74/196, 37.8%) and ganciclovir (67/196, 34.2%) were almost equally used in HCT recipients. Ganciclovir, a nucleoside analogue, is likely preferred in SOT as it has fewer nephrotoxic effects than foscarnet. Foscarnet also causes multiple electrolyte imbalances and QTc prolongation, which poses increased difficulty in terms of monitoring, particularly among kidney allografts. In contrast, ganciclovir and valganciclovir are generally avoided in HCT recipients due to their myelosuppressive effects, which can delay marrow recovery. Unfortunately, the optimal drug of choice and the duration of therapy against HHV-6 are unknown and need to be defined. To highlight the key features of HHV-6, a summary is provided ( Table 5 ).

Although the attributable mortality from HHV-6 was modest (22/107, 20.6%), at least a third of patients (45/151, 29.8%) had some neurological deficits at the last follow-up, long after resolution of the clinical disease. The incidence may even be higher, as the neurological outcome was unspecified in a fourth of cases (38/151, 25.2%). Morbidity from HHV-6 disease can be debilitating; as such, HHV-6 neurological involvement should be recognized early and treated aggressively.

Other unique observations from this systematic review deserve mention. Firstly, the higher proportion of men could suggest that they are disproportionately affected by HHV-6. Although this may be a reflection of the overall gender disparity of the transplant population, which is predominantly male, this merits further exploration. Secondly, CMV co-infection was more frequently reported in hematopoietic than in SOT recipients, which mirrors the expectations in the larger transplant population. This review also encompasses a period where CMV surveillance and preemptive therapy comprise a more common CMV prevention strategy in hematopoietic stem cell transplants compared with universal prophylaxis among SOT recipients. However, most case reports did not explicitly state whether CMV was evaluated, and the risk of CMV reactivation may have been underreported overall. Finally, HHV-6A appears to be as prevalent as HHV-6B among SOT recipients, which challenges the notion that HHV-6B is more common in the transplant population. However, as the numbers are limited, no definitive conclusions can be made, and this must be interpreted with caution.

Our study has several limitations. Firstly, there are duplications across studies (e.g., case reports and cohorts) as some cases from larger cohorts were described in detail. Secondly, we excluded asymptomatic HHV-6 and ciHHV-6, but these could have been inadvertently misclassified as asymptomatic when true disease was present. Alternatively, other viral infections (e.g., CMV and EBV) or syndromes (e.g., ICANS) can present similarly, and symptoms attributable to HHV-6 could have been from these other causes. Important publications in non-English journals may have been missed as the language was limited to English. Reports lacking patient-level data, pediatric transplants, and abstracts of cohort studies were also excluded; thus, our numbers likely underestimated the true magnitude of HHV-6. Finally, due to the heterogeneity of cohort and case–control studies, we were unable to perform a meta-analysis. A formal assessment of risk factors predictive of HHV-6 disease, for example, may be helpful for transplant clinicians, and a future analysis needs to be performed. Despite these, our study is the first to provide cumulative data on HHV-6 for transplant and CAR-T recipients.

Our review highlights that HHV-6 subclinical reactivation frequently occurs. HHV-6 clinical disease is more commonly reported among HCT compared with SOT or CAR-T recipients. However, only six CAR-T cases and one cohort study were included, prohibiting conclusions for this group. Neurological disease in the form of encephalitis is the most frequent clinical syndrome. Early recognition of limbic involvement either through the triad of confusion, amnesia, and seizures or through compatible MRI findings may help with the early identification of HHV-6. Diagnosis is secured through molecular methods, although HHV-6 detection per se may not necessarily indicate disease. Extremely high viremia needs to be interpreted with caution in the context of ciHHV-6, which may not warrant treatment. HHV-6-associated mortality is modest; however, the neurological sequelae can be disabling and cause significant morbidity.