HHV-6-Associated Post-Transplantation Acute Limbic Encephalitis Following Cord Blood Stem Cell Transplantation: A Cohort Analysis

Abstract 649

Human Herpesvirus-6 (HHV6) frequently reactivates after allogeneic hematopoietic stem cell transplantation (HSCT). One serious manifestation of HHV6 reactivation is the syndrome of post-transplantation acute limbic encephalitis (PALE) associated with HHV6 reactivation in the central nervous system. We previously described this disease after peripheral blood HSCT (PBSCT), but the epidemiology and characteristics of PALE in patients receiving unrelated cord blood transplantation (UCBT) is not well characterized.

We analyzed all patients undergoing allogeneic HSCT at our center from 3/2003 through 3/2010. HHV-6-associated PALE (HHV6-PALE) was diagnosed in patients who had a positive CSF PCR test for HHV6 DNA in the context of a) acute-onset altered mental status, anterograde amnesia, or seizures and/or b) MRI or EEG abnormalities involving the limbic system with no other identifiable etiology. Medical records were reviewed for demographic and HSCT characteristics, as well as the clinical, laboratory and radiographic features of cases. Patient time-at-risk was censored at day of HHV6-PALE symptom onset, death or day +100 after HSCT, as all cases occurred within this period. Incidence rates were calculated for pertinent characteristics, and Cox modeling was used to analyze potential HHV6-PALE risk factors.

A total of 1,344 patients underwent an initial HSCT during the study period: 725 from adult unrelated donors, 518 from adult related donors and 101 from UCB donors. Fifteen patients underwent a second HSCT procedure (8 UCB and 7 PBSC) during the 100-day follow up period. HHV6-PALE was diagnosed in 19 patients. Two of these cases occurred after a subsequent UCBT. The cumulative incidence of HHV6-PALE was 1.4% for an overall incidence rate of 0.15/1000 patient-days (95% confidence interval (CI), 0.09–0.24). HHV6-PALE incidence rate was higher among UCBT patients (10/101, IR 1.2/1000 patient-days) compared with the rest of the cohort (9/1243, IR 0.08/1000 patient-days, p<0.001). Other relevant characteristics associated with HHV6-PALE included acute graft-versus-host disease (GVHD) grade II-IV (p=0.05), adult mismatched donor (p=0.03) and conditioning with thymoglobulin (p=0.003). On multivariable Cox modeling, UCBT (adjusted HR 20.0, 95% CI, 7.3–55.0), adult mismatched donor (adjusted HR 4.3, 95% CI, 1.1–17.3) and time-dependent acute GVHD (adjusted HR 7.5, 95% CI, 2.8–19.8) remained significant.

Interesting clinical features were noted when comparing the 10 UCBT HHV6-PALE cases to the 9 PBSCT cases. Encephalitis developed prior to engraftment in 7 recipients of UCB compared with 1 recipient of PBSC. HHV6-PALE symptom onset occurred at a median of 20 days (range, 7–37) in PBSCT compared with 32 days (range, 16–67) in UCBT patients (p=0.07). Brain MRI abnormalities were limited to the limbic system in PBSCT recipients but extended beyond the temporal lobes in 2 UCBT patients. Intravenous foscarnet was used to treat 18/19 patients for a median of 21 days (range, 7–42); time to treatment after symptom onset was a median of 3 days (range, 1–13) in PBSCT and 6 days (range, 1–13) in UCBT patients. No PBSCT patients died from HHV6-PALE, although most patients had long-term neurocognitive deficits. Five UCBT patients died a median of 45 days after transplant and 18 days after symptom onset. Deaths occurred after similar courses punctuated by progressive encephalopathy and unresponsiveness requiring mechanical ventilation. Of 68 UCBT patients who underwent plasma HHV6 PCR testing, 49 (72.1%) had positive results. All patients with HHV6-PALE had detectable HHV6 DNA in their plasma. HHV6 PCR results were higher in patients with HHV-6 PALE (median 173,500 copies/mL, range, 7,100– >10⁶) than in patients without HHV6-PALE (median 8,160 copies/mL, range, <1,000– >10⁶, p=0.003). Plasma HHV6 PCR values greater than 250,000 copies/mL were 95% specific for a diagnosis of HHV6-PALE.

Patients receiving UCBT are at increased risk for developing HHV6-PALE. This disease has different characteristics after UCBT with greater morbidity and mortality. Preventive strategies to minimize the impact of HHV6-PALE in this population need to be further evaluated.