Tomoko Teramura

Requirement for Etoposide in the Treatment of Epstein-Barr Virus–Associated Hemophagocytic Lymphohistiocytosis

PURPOSE We sought to identify the clinical variables most critical to successful treatment of Epstein-Barr virus (EBV)-associated hemophagocytic lymphohistiocytosis (HLH). PATIENTS AND METHODS Among the factors tested were age at diagnosis (< 2 years or > or = 2 years), time from diagnosis to initiation of treatment with or without etoposide-containing regimens, timing of cyclosporin A (CSA) administration during induction therapy, and the presence or absence of etoposide. RESULTS By Kaplan-Meier analysis, the overall survival rate for the entire cohort of 47 patients, most of whom had moderately severe to severe disease, was 78.3% +/- 6.7% (SE) at 4 years. The probability of long-term survival was significantly higher when etoposide treatment was begun less than 4 weeks from diagnosis (90.2% +/- 6.9% v 56.5% +/- 12.6% for patients receiving this agent later or not at all; P <.01, log-rank test). Multivariate analysis with the Cox proportional hazards model demonstrated the independent prognostic significance of a short interval from EBV-HLH diagnosis to etoposide administration (relative risk of death for patients lacking this feature, 14.1; 95% confidence interval, 1.16 to 166.7; P =.04). None of the competing variables analyzed had significant predictive strength in the Cox model. However, concomitant use of CSA with etoposide in a subset of patients appears to have prevented serious complications from neutropenia during the first year of treatment. CONCLUSION We conclude that early administration of etoposide, preferably with CSA, is the treatment of choice for patients with EBV-HLH.

Occurrence of haemophagocytic lymphohistiocytosis at less than 1 year of age: analysis of 96 patients

We analysed data of 96 infants (under 1 year of age) with haemophagocytic lymphohistiocytosis (HLH) from the registry of an HLH study conducted during 1986–2002 in Japan. The cases were classified into five groups. The diagnosis of familial HLH (FHL) as group 1 (n =27) was made with positive family history and/or recent molecular test for perforin and Munc13–4 mutations. Neonatal enterovirus- or herpes simplex virus-associated HLH as group 2a (n =7), Epstein-Barr virus–associated HLH (n =12) as group 2b, adenovirus- or cytomegalovirus-associated HLH as group 3 (n =9) were mostly diagnosed by viral isolation or by the detection of viral genome. Juvenile rheumatoid arthritis-associated macrophage activation syndrome was classified as group 4 (n =4) and the remaining without known triggers as group 5 (n =37). The peak onset age was 1–2 months for group 1, 1–2 weeks for group 2a, 12 months for group 2b, none for group 3, 9 months for group 4 and 2 months for group 5. Future novel diagnostic measures are required to define the precise nature of HLH in group 5. Conclusion:These data may provide useful information for neonatologists/ paediatricians in the differential diagnosis of haemophagocytic lymphohistiocytosis in early infancy.

Quantitative Analysis of Cell-free Epstein-Barr Virus Genome Copy Number in Patients with EBV-associated Hemophagocytic Lymphohistiocytosis

To determine whether the EBV genome content in serum or plasma reflects clinical features and outcome in EBV-associated hemophagocytic lymphohistiocytosis (EBV-HLH), we quantified the cell-free EBV genome copy number by real-time PCR in 38 patients with EBV-HLH, and compared this to the values from 15 patients with infectious mononucleosis (IM). The median (range) cell-free EBV genome copy number at diagnosis was 3.0   ×   10 3 (undetectable m 5.5   ×   10 7 ) copies/ml in EBV-HLH, which was significantly higher than the 6.6   ×   10 1 (undetectable m 1.0   ×   10 3 ) copies/ml in IM (P =0.0008). We serially analyzed cell-free EBV genome copy number in 10 cases of EBV-HLH up to 4 months from diagnosis. In four patients who achieved remission, the EBV genome became undetectable soon after starting therapy. In the remaining six patients who responded poorly to therapy, the EBV genome copy number in the serum or plasma remained at high levels except for one case. In addition, we confirmed that the EBV genome became undetectable after hematopoietic stem cell transplantation in 4 EBV-HLH cases. These results suggest that the quantitative analysis of cell-free EBV genome copy number is useful for evaluating disease activity and for predicting the response to therapy in EBV-HLH.

Treatment Strategies for Epstein-Barr Virus-Associated Hemophagocytic Lymphohistiocytosis (EBV-HLH)

In Epstein-Barr virus (EBV) infection, the virus immortalizes B lymphocytes and cytotoxic T lymphocytes (CTLs) are directed toward both latent and lytic viral antigens expressed on EBV-infected B-cells. Various EBV-associated diseases occur as a result of this disruption of immune surveillance. In the majority of EBV-associated hemophagocytic lymphohistiocytosis (EBV-HLH) cases, the major cell types containing EBV DNA are not B-cells, but clonally proliferating T-cells or NK-cells. Proliferation of these cells produces severe immune reactions in the host, and the clinical features related to massive cytokine production at the onset of disease are unique and distinct from other EBV-associated diseases. In the treatment of EBV-HLH, therapeutic infusion of EBV-specific CTLs appears to be ineffective, and eradication of EBV-containing cells is useful but not sufficient to save lives, because of high incidence of acute mortality due to cytokine-induced multiple organ failure and neutropenia-associated opportunistic infections. The optimal treatment strategy for this disease consists of three steps: (1) control of cytokine storm including coagulopathy and multiple organ failure, (2) control of opportunistic infections, and (3) eradication of clonally proliferating EBV-containing T- or NK- cells by immunochemotherapy and, if necessary, hemopoietic stem cell/ bone marrow transplantation (SCT/BMT).

Recent developments in the management of haemophagocytic lymphohistiocytosis

Over the past two decades, the underlying pathophysiology of haemophagocytic lymphohistiocytosis (HLH) (synonyms: haemophagocytic syndrome, macrophage activation syndrome) has been well recognised. Cytokine storm plays a major role, which derives from an inappropriate immune reaction caused by proliferating and activated T-cell or natural killer (NK) cells associated with macrophage activation and inadequate apoptosis of immunogenic cells. Many biological parameters reflecting activity of disease or response to treatment have been identified, in particular, serum ferritin has been confirmed to be one of the markers for HLH. The common types of HLH consist of non-hereditary (acquired) infection-associated disease such as Epstein-Barr virus (EBV)-haemophagocytic lymphohistiocytosis (HLH) and hereditary (familial) disease such as FHL, in which, at the molecular level, dysfunctional perforin was clarified. Regarding the therapeutic strategies, prompt differential diagnosis of underlying disease is essential and choice of treatment should be based on the risk (low or high) of prognosis, where either cyclosporin A, steroids or iv. immunoglobulin (IVIG) may be indicated as initial treatment for low-risk patients, with etoposide-containing regimens for high-risk patients. Significant improvement of prognosis has been obtained by incorporating intensive supportive care at the disease onset and prompt introduction of immunosuppressants to control cytokine storm. Subsequent immunochemotherapy and haemopoietic stem cell transplantation have contributed significantly to further improve survival of hereditary and refractory HLH patients.

Risk of Etoposide-Related Acute Myeloid Leukemia in the Treatment of Epstein-Barr Virus—Associated Hemophagocytic Lymphohistiocytosis

We studied the impact of etoposide on the prognosis of 81 patients (77 of whom were children <15 years old) with Epstein-Barr virus—associated hemophagocytic lymphohistiocytosis (EBV-HLH).The study group received a median cumulative dose of 1500 mg/m2 etoposide (range, 0–14,550 mg/m2), with a median follow-up period of 44 months (range, 20–88 months) from the diagnosis. Only 1 patient, who received 3150 mg/m2 etoposide, developed therapy-related acute myeloid leukemia (t-AML), at 31 months after diagnosis. Excluding 9 patients who underwent hemopoietic stem cell transplantation during the course of treatment, the prognosis was poorer for those patients who received less than a 1000 mg/m2 cumulative dose of etoposide. Our results indicate that the risk of etoposide-related t-AML is low. An appropriate dosage of etoposide for the treatment of EBV-HLH would be in the range of 1000 to 3000 mg/m2. However, even at these doses, care must be taken to prevent the rare risk of t-AML.

Quantitative detection of serum adenovirus in a transplant recipient

Sir—Adenovirus infection in compromised hosts can become a fatal infection. This disorder, especially adenovirus-triggered severe haemorrhagic cystitis and systemic disease, is one of the most troublesome complications in transplant recipients, and has high morbidity. Serotypes 11, 34, and 35 are most common. We have established a real-time PCR system following the previously described procedure for adenovirus serotype 35. The PCR uses primers upstream 5 -CAGGTAGACTGCCT CGATGATG-3 , downstream 5 -GC CCACCCTGCTTTATCTTCTC-3 , and with a probe of 5 -(FAM)TGCA CTCTGACCACGTCGAAAACTTC (TAMRA)-3 cutoff 50 copies/mL in sera and 50 copies/ g creatinine in urine). We monitored serum and urine adenovirus genome quantitatively, although retrospectively, in one transplant recipient with severe adenovirus systemic disease. The viral genome load well reflected the disease activity. A girl aged 10 years with acute myeloblastic leukaemia (FAB, M3) underwent bone-marrow transplantation from a matched sibling donor after haematological relapse, with busulfan, cyclophosphamide, and melphalan conditioning, and with prophylaxis for graft-versus-host disease of cyclosporin A, methotrexate, and prednisolone. The transplant was successful, but the patient developed haemorrhagic cystitis starting at day 25, followed by severe systemic infection with high fever. We isolated adenovirus serotype 35 from her urine on day 26. Our PCR system showed that samples before transplantation were negative. Serum adenovirus genome became detectable on day 15 with around 75 10 copies/mL. As shown in the figure, serum adenovirus genome increased up to 2·5 10 copies/mL on day 60. Mild decrease of adenovirus genome might have reflected the treatment effect of intravenous vidarabin and ribavirin. Because of bone marrow failure, probably due to systemic adenovirus infection and adverse effect of antiviral treatment, a second bone-marrow transplant from the same donor was done on day 110. Even after the second transplantation, serum adenovirus genome remained high at 5 10 copies/mL, clinically associated with persistent fever and microhaematuria. With intensive care, the patient survived with late sequelae of renal and cardiac dysfunction. The adenovirus genome in urine was similarly changed as those in sera (figure). The detection of serum genome by PCR could be the first step towards predicting the outlook in transplant recipients with adenovirus infection. We suggest that real-time PCR is also useful for assessment of treatment response, as shown in post-transplant Epstein-Barr virus infection.

Molecular Analysis of Latent Membrane Protein 1 in Patients with Epstein-Barr Virus-Associated Hemophagocytic Lymphohistiocytosis in Japan

Latent membrane protein 1 (LMP1) of Epstein-Barr virus (EBV) is considered to be an oncoprotein because it is crucial for B-lymphocyte transformation. Since a 30 base pair (bp) deletion in the carboxy-terminal portion of the LMP1 gene was found in a CAO cell line derived from nasopharyngeal carcinoma containing EBV, an association between EB viral genetic alteration and tumorigenicity has been postulated. In this study we have analyzed LMP1 DNA isolated from 10 Japanese patients with EBV-associated hemophagocytic lymphohistiocytosis (EBV-HLH). In all HLH patients, we found the 30bp deletion and 4-8-tandem repeats of the sequence DNGPQDPDNTD in the LMP1 gene. Furthermore, detailed amino acid (aa) sequence analysis revealed that 7 aa substitutions identical to those found in CAO-LMP1 but not in B95.8 cell line-LMPl were found in all the HLH cases. NF-κB assay revealed that HLH-LMP1 activated NF-κB significantly more than that of B95.8-LMP1 (p=0.032). We conclude that EBV from all of our HLH cases shared common genetic characteristics with EBV obtained from the CAO cell line, which is distinct from the wild-type EBV isolated from the B95.8 cell line. These data suggest that the mutational changes of the LMP1 gene may play an important role in the pathogenesis of these fatal EBV-related disorders.

Hemophagocytic Syndrome Associated with Severe Adenoviral Pneumonia: Usefulness of Real-Time Polymerase Chain Reaction for Diagnosis

In infection-associated hemophagocytic syndrome (HPS) the causative pathogen is often undetected, except in cases of herpes virus infections. We describe a 12-year-old girl with life-threatening pneumonia with HPS caused by an adenovirus. She was admitted with complaints of persistent fever and systemic petechiae/purpura.The day after admission the patient developed sudden dyspnea with massive infiltration of the bilateral lower lung field. She exhibited coagulopathy, hepatic dysfunction, and remarkable elevations in serum levels of cytokine, ferritin, and urinary β2-microglobulin.A diagnosis of HPS was made, and the patient was treated with dexamethasone and cyclosporin A on the second hospital day. Her fever went down quickly, and the abnormal laboratory and chest radiographic findings returned to normal over a period of 2 weeks.Antibody analysis was not successful in identifying the pathogen responsible.However, a polymerase chain reaction (PCR) assay of lung tissue biopsied on the fifth hospital day was positive for an adenovirus (subgroup B), and quantitative adenoviral DNAanalysis by real-time PCR using primers covering serotypes 3, 7, 11, and 35 (all subgroup B) confirmed this initial finding (93 copies/μg DNA in the biopsied lung and no detectable adenovirusDNAin the lung tissues of control subjects).This approach may provide important clues for improving outcomes and clarifying the exact etiology in cases of often fatal, infection-associated HPS.

Sensorineural hearing loss in a case of familial hemophagocytic lymphohistiocytosis

Severe sensorineural hearing loss (bilateral >80 dB) was diagnosed in a case of familial hemophagocytic lymphohistiocytosis (FHL). The female patient developed HLH at 3 months of age and underwent allogeneic cord blood transplantation at 11 months of age following 7 months of immuno‐chemotherapy. The type 2 FHL patient had a homozygous perforin gene mutation of 1090–1091delCT, and was noted to have hearing loss at 3.5 years of age. Retrospective evaluation did not clarify the exact causes of hearing loss. Reports on Kawasaki disease, suggesting a correlation between severe inflammatory status in infancy and the development of sensorineural hearing loss, may shed some light on this rare complication in this case of FHL. Considering the markedly improved prognosis of FHL due to recent advances made in the molecular diagnosis and in the management including allogeneic hematopoietic stem cell transplantation, auditor by screening might be warranted for surviving FHL patients. Pediatr Blood Cancer 2007;49:856–858.