Background
Hemophagocytic Syndrome (HPS), also known as hemophagocytic lymphohistiocytosis (HLH), is a potentially life-threatening immune system disorder characterized by cytokine storm and overwhelming inflammation [
1]. Cytotoxic cells and macrophages cause multiorgan damage, hemophagocytosis, and severe systemic inflammation [
2]. The clinical presentations of HLH were generally prolonged fever, hepatosplenomegaly, cytopenia, hypertriglyceridemia, hyperferritinemia, and hemophagocytosis in bone marrow, liver, spleen or lymph nodes [
3]. Patients with primary HLH usually have a family history of the disease, or known underlying genetic defects which predispose them to the disease [
4]. Secondary HLH is usually caused by some etiologies including infections, autoimmune diseases, malignancies, acquired immune deficiency, as well as iatrogenic immune suppression and organ or stem cell transplantation [
5,
6].
HLH is a relatively rare disease but has garnered increased attention over the past ten years. More than 1500 publications on this topic have appeared since 2004. Moreover, HLH was once considered a children’s disease (<14 years old). However, cases are being reported with increasing frequency in adults [
7]. Patients with HLH present with a wide spectrum of clinical manifestations, rapidly deteriorating conditions, as well as considerable morbidity and mortality.
To improve the recognition and understanding of this disorder in elderly HLH patients (>14 years old) in China, we analyzed data from 85 patients diagnosed at the First Affiliated Hospital of Nanchang University from April 2005 to June 2014. We demonstrated the variable clinical spectrums of these patients, and determined the poor prognosis factors related to HLH patient’s survival.
Discussion
HLH is an infrequent but potentially life-threatening hyperinflammation syndrome, which is caused by a range of inherited or secondary factors. The onset of F-HLH is usually within the first 6 months of life, however, patients with a later onset, even up to adulthood have also been reported [
13,
14]. Five different forms of F-HLH have so far been described and four genes (
PFR1,
UNC13D,
STX11,
STXBP2) have been identified. Secondary HLH (sHLH) can develop at any age and is usually caused by underlying conditions such as infection, autoimmune/rheumatologic, malignant conditions et cetera [
1]. Currently, data as reported in the literature is not precise enough to predict the incidence of HLH, although single-center studies have reported the incidence to be 0.12/100,000 per year in familial-HLH (F-HLH) patients and 0.36/100,000 per year in M-HLH patients [
15]. HLH in young children/infants, mainly occurring F-HLH with a genetic defect, developed similar symptoms to patients from our study with a couple exceptions. Similar symptoms included fever (91%), splenomegaly (81%), hepatomegaly (81%), and lymph node enlargement (43%), however, the incidence of neurologic symptoms (47%), and Rash (43%) might be higher in F-HLH and neurologic symptoms occurred earlier than other symptoms according to Zhang’s report [
14]. This study has focused on the clinical characteristics and prognostic factors of adult HLH patients to improve the recognition and understanding of this disorder.
First manifestation of HLH could resemble common infections, malignancies, fever of unknown origin, or autoimmune disorders [
16]. The clinical features are thought to be due to hypercytokinemia, such as tumor necrosis factor TNF-α, interferon (IFN)-γ, interleukin (IL)-10, IL-12, and IL-18 released by highly activated lymphocytes and macrophages [
17]. Typical clinical findings are prolonged fever, insensitivity to antibiotic therapy, and hepatosplenomegaly. Laboratory findings include cytopenias, usually beginning with thrombocytopenia evolving into severe pancytopenia, hyperferritinemia, elevated transaminases, hypofibrinogenemia, DIC, hypertriglyceridemia, hypoalbuminemia, and hyponatremia [
18]. Our data also showed that patients with HLH had variable clinical spectrums as mentioned above (Figure
1A). Due to the confusing clinical symptoms, 49% of the patients were initially misdiagnosed and hospitalized in infectious, gastroenterology, respiratory, or other departments for fever, liver lesion, dizziness, or gastrointestinal bleeding. Due to our increasing recognition based on clinical and laboratory report and the improvements in pathological diagnosis techniques in our hospital, the numbers of patients diagnosed as HLH raised significantly in our institution. In addition, since deficient NK-cells activity and sCD25 are hallmark of HLH and exhibited in major HLH patients [
19], the lacking these two tests may veil the real HLH patients numbers diagnosed in our institution. Therefore, we infer that the incidence of HLH occur more commonly than it appeared at least in Jiangxi state.
In this study, we queried the most common form of HLH in adults, infection associated HLH. Infectious triggers include viruses (for example, EBV, cytomegalovirus, HIV), bacteria (for example, mycobacteria), and fungi (for example, candida, cryptococcus) [
20]. Among viral infections, EBV is undoubtedly the major cause of HLH. Published data have shown that very high levels of proinflammatory cytokines are associated with EBV-related HLH among Asians [
21]. In our current study, 33.8% of patients had EBV infection which is consistent with the result of a Japanese study (33.3%) [
22]. Interestingly, a study showed a quarter of male patients with EBV-associated HLH may have mutations in the
SH2D1A gene, which is traditionally associated with X-linked lymphoproliferative syndrome (XLPS) immunodeficient to EBV [
23]. In our data, male HLH patients presented with EBV infections in 36.1% of cases. In contrast, female patients had a rate of 23.8%, moreover, the incidence of HLH was higher in male patients than in females (3:1 ratio). Determining if the higher HLH incidence in males was related to EBV infections will need to be confirmed in a future study. The clinical outcome heterogeneity of patients with EBV-HLH is striking, ranging from self-limiting to aggressive and fatal. Some results indicated that patients with active EBV-HLH or high EBV genome copy numbers had poor prognosis [
24,
25]. In the infection subgroup, we found patients with active EBV infection had shorter survival than non-EBV infected patients (65d vs. not reached,
P = 0.021) but there was no significance in the M-HLH subgroup. The high mortality rate among patients with EBV-HLH is due to, at least in part, delays in diagnosis that result from the similarity between its initial clinical presentation and common infective and inflammatory conditions. Of note, improving clinical outcomes and diagnosing patients with refractory EBV-HLH is still challenging. However, outcomes have improved significantly with the use of rituximab, etoposide, and HCT for these patients [
26].
We continued to investigate malignant manifestation in HLH patients. M-HLH develops most frequently in patients with T and NK-cell lymphoma, the strong association between the T-cell and NK-cell lymphomas and HLH was reported in several previous studies [
27,
28]. T and NK-cell associated HLH has also been shown to have the lowest five year survival rate (12%) [
14]. Proper T and NK-cell function is required for clearance of antigenic stimuli and termination of the inflammatory response. Aberrant T-cell and NK-cell activation results in excessive cytokine production and sustained macrophage activation. Clonality studies have shown that a significant number of patients with EBV-associated HLH have a clonal proliferation of T cells [
29,
30]. M-HLH can also occur in the course of other hematological malignancies (e.g., B-cell lymphoma, Hodgkin’s lymphoma, acute and chronic leukemias) [
31]. Ishii et al. [
32] reported that in Japan the causes of 18% of M-HLH cases were cancers other than lymphoma such as acute myeloid leukemia and myelodysplastic syndromes. Our results were in accordance with the above, 82.6% of the patients in this subset (N = 23) had T and NK cell lymphoma/leukemia, three patients had B-cell lymphoma, and one patient had T lymphoblastic leukemia. 47.8% (11/23) of patients simultaneously had EBV infection.
HLH patients have a high mortality rate, which is partially caused by a lack of clinical suspicion and unavailability of genetic as well as other molecular studies in most developing countries. Therefore, it is necessary to establish effective pretreatment markers to predict the high-risk subsets in this disorder. In this study, we further explored the prognostic factors related to HLH survival in clinics. Our results showed patients with M-HLH had the worst prognosis (median OS: 40 days), and patients in unexplained cause group also had poor outcomes (median OS: 90 days. This may have been caused in part by an inability to diagnose lymphoma in patients that were suspected of having it due to lack of evidence.
To exclude the interference of different hospitalized divisions, we compared the survival time among patients from various departments. The results revealed that patients initially hospitalized in hematology had the shortest survival time (median OS, 45 days), followed by patients hospitalized in other departments (median OS, 350 days) and infectious department (median OS, not reached) (
P = 0.006) (data not shown). This result is consistent with our conclusion that M-HLH and EBV infection are associated with poor survival. This is due to the fact that there were 19 patients with lymphoma and 10 with EBV infection admitted to the hematology department. Also, there was no survival difference between patients diagnosed before and after 2011 (data not shown). Intriguingly, we found patients with Fbg <1.5 g/L, PLT <40 × 10
9/L, or LDH ≥1000 U/L had shorter OS by using univariate analysis which is consistent with our previous study in lymphoma - associated HLH (LAHS) [
33]. Furthermore, multivariate analysis demonstrated PLT <40 × 10
9/L was the sole predictor of poor OS, which is consistent with the reports from Wang Z et al. [
34,
35]. The mechanism under which PLT is operating remains elusive. We inferred that PLT change might be a more direct consequence of cytokine storm and hyperinflammation compared to Fbg and LDH.
HLH is an aggressive and incurable disease. Even with treatment and haematopoietic stem cell transplant (HSCT), the 5-year survival probability ranges from 45% to 75%, but the survival was 0% without HSCT treatment [
12]. Due to the aggressive life-threatening implications of HLH, effective treatments including therapies that target activated macrophages/histiocytes (etoposide, steroids, high-dose IVIgG) and/or activated T cells (steroids, cyclosporine A, antithymocyte globulins) [
36] should be instituted promptly, followed by HSCT. The survival of patients diagnosed in a study conducted at South Carolina University with EBV-HLH was significantly improved when etoposide treatment started within 4 weeks of diagnosis [
37]. Besides, treatment on coexisting infections, identification of other potential triggers of HLH is also necessary. Patients who do not respond to steroids require management with aggressive combination chemotherapy, followed by HSCT. In our group, the median OS is 60 days despite the implementation of treatments such as etoposide, glucocorticoid and cyclosporine. The reason for poor survival in these patients may be as follows, the unavailability of some high specificity indicator tests [
38,
39] such as soluble CD25 levels and NK cell activity in our institution resulted in delayed diagnoses and treatment of HLH patients which led to end-organ failure and death. Moreover, high-dose chemotherapy drugs including etoposide and allo-HCT regimens were not applied in these patients. These treatments could considerably improve survival for patients with refractory EBV-associated HLH and lymphoma [
12].
In this article, our results echo clinical studies completed elsewhere that both T/NK-cell abnormalities as well as EBV viral infection play an important role in the disease progression [
27,
30,
32]. This study attempted to evaluate effective markers for diagnoses across a wider spectrum of cases and long follow-up time (up to 9 years). Studies in the past usually focused on identifying sole clinical marker to this disease [
37]. With a better idea of confidence and significance, we disclose the importance of factors like malignance disease and EBV infection et al. simultaneously on the sHLH based on a comprehensive interpretation of solid sample. However, other studies showed some different prognostic factors such as the high level of serum ferritin to our data [
35]. More studies would need to be performed to verify these results.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
FL designed, analyzed the data, and wrote the paper. FJ, JR, PL, YZ, GY contributed to clinical data collection. YY, RZ and MW contributed to statistical analysis and discussed the results. YY and CD contributed to the revision of the manuscript, MW, RZ, YY and FL contributed to the approval of the final manuscript. All authors read and approved the final manuscript.