Treatment of Familial Hemophagocytic Lymphohistiocytosis with Third-Party Mesenchymal Stromal Cells

Introduction

M esenchymal stromal cells (MSCs) are a heterogeneous group of fibroblast-like, multipotent cells [1,2]. They can be isolated from different tissues, including bone marrow, umbilical cord, and fat. Beyond their regenerative properties, MSCs display a plethora of immunomodulating mechanisms. They suppress T- and NK-cell responses, and repolarize proinflammatory antigen-presenting dendritic cells (DCs) and macrophages into tolerogenic phenotypes. Furthermore, MSCs are considered immunoprivileged, which allows a third-party cell transfer without eliciting immune rejection mechanisms. Based on these features, MSCs have emerged as an attractive cell therapeutic tool for inflammatory and autoimmune diseases. Multiple clinical trials have already indicated beneficial effects of MSC infusion for treatment of acute graft-versus-host disease (GVHD) as well as Crohn's disease [3,4], whereas no acute toxicity or increased risk for infectious complications has been reported [2,5]. The familial hemophagocytic lymphohistiocytosis (FHL) is an immunodeficiency disorder with autosomal recessive inheritance and a median survival of 2 months, if untreated [6,7]. FHL is characterized by a pronounced systemic hyperinflammation that results from a defect in the lymphocytes' cytotoxic function. It is associated with mutations in a variety of genes responsible for cytotoxicity, including PRF1 and UNC13D [8]. The onset of the disease can be triggered by infections, when pathogen clearance as well as crosstalk between innate and adaptive immunity are impaired, leading to hyperactivation and hyperproliferation of cytotoxic T-cells and macrophages. The clinical presentation typically includes fever, hepatosplenomegaly, and cytopenias. Other diagnostic criteria are hypertriglyceridemia, hemophagocytosis (in bone marrow, lymph nodes, or spleen), reduced NK-cell activity, hyperferritinemia, and increased soluble IL-2 receptor (CD25) levels. Treatment is based on the combination of chemotherapy and immunomodulation using etoposide, cyclosporine, and steroids (HLH-94 and HLH-2004 protocols) [7,9]. This strategy is moderately effective in the control of hyperinflammation, but the outcome is usually fatal unless the patient undergoes allogeneic stem cell transplantation (allo-SCT) [6,7,9].

Promising observations in severe inflammatory conditions such as the acute GVHD and the key role of immune alterations in FHL make MSCs a rational therapeutic approach in FHL. Here, we present our experience on the first use of MSCs in a patient with otherwise treatment-resistant FHL.

Design and Methods

Results and Discussion

In October 2010, a 21-year-old previously healthy male of mixed ethnic background (half Swedish and half Chinese) was referred to the Karolinska University Hospital with a 2-month history of persistent fever (39.5°C) and pancytopenia that had started after he travelled back to Sweden from Asia. The clinical picture was interpreted as refractory infection-associated hemophagocytic lymphohistiocytosis (HLH) (8/8 diagnostic criteria [6]) triggered by gastrointestinal infections with Salmonella spp. and Entamoeba spp. that had been successfully eradicated. On admission, the patient was administered betamethasone and intravenous immunoglobulin that resulted in a prompt, but temporary (1-week) response. Therapy was changed to a modified HLH-94 immunochemotherapy, with immediate clinical improvement (eg, normalized body temperature and reduced levels of ferritin and soluble CD25). After 2 months, his HLH recurred and was difficult to control, as chemoimmunotherapy was only marginally effective in improving clinical (fever and hepatosplenomegaly) and laboratory parameters (Fig. 1). Analysis of the NK-cell activity indicated severely hypofunctional NK-cells typical of FHL (data not shown). Sequencing revealed compound heterozygous mutations (c.2039_2040GG>TT; p.Arg680Leu and c.2296C>T; p.Gln766×) in the UNC13D gene encoding the Munc 13-4 protein, establishing the diagnosis of a late-onset FHL type 3 [8]. Munc 13-4 is required for the maturation and secretion of cytotoxic granules by cytotoxic NK-cells and T-cells [11].

OPEN IN VIEWER

FIG. 1.

Treatment and laboratory values before and after infusion with MSCs. Abbreviations: WBC, white blood cells; TG, triglycerides; LDH, lactate dehydrogenase; TNFa, tumor necrosis factor a; MSC, mesenchymal stromal cell. Reference values: platelets 145–348×109/L; WBC 3.5–8.8×109/L; ferritin 30–350 mg/L; triglycerides 0.45–2.6 mmol/L; LDH<210 U/L; D-Dimer<0.25 mg/L; IL-6<5 pg/mL; IL-8<60 pg/mL; TNFa<12 pg/mL. Color images available online at www.liebertpub.com/scd

A search for an HLA-matched unrelated donor for allo-SCT was initiated, but the clinical status of the patient deteriorated rapidly despite a continued HLH-94 protocol. Salvage therapies for HLH (eg, Campath^®|Alemtuzumab; a monoclonal antibody that targets CD52 expressed on activated lymphocytes and macrophages [12]) in conjunction with the disease-related immune dysfunction pose a substantial risk for infections in HLH patients. In contrast, MSC therapy has so far not been associated with an increased risk of infection even in severely immunocompromised patients such as in GVHD [13]. In fact, increasing evidence suggests that MSCs might even have a role in the host defense [2,5]. After a careful consideration, we decided to give the patient one dose of MSCs. T-cell-suppressive bone marrow-derived MSCs (1.4×10⁶ cells/kg of body weight, in total 124×10⁶ cells) from an unrelated healthy donor (Fig. 2B) were given intravenously (day 0). Within 24 h, the patient was afebrile (<37.5°C body temperature), and his clinical condition and laboratory parameters (especially ferritin, triglycerides, and lactate dehydrogenase) quickly improved (Fig. 1). Chemotherapy was discontinued. Concomitantly, levels of several proinflammatory cytokines and chemokines became reduced together with rising levels of immunosuppressive IL-10 (Figs. 1 and 2C). Due to a Rhizopus microsporus infection that was present before MSC infusion, the patient underwent surgical resection of the left orbita and surrounding sinuses on day +8 (Fig. 1). Two days later, the fungal infection had become disseminated, and the patient died on day +13 from severe fungal sepsis. In the following autopsy, MSC donor DNA was only detectable at very low levels in the heart tissue, as we have published elsewhere [14], and which is in line with the proposed hit-and-run-like effects mediated by MSCs without sustained engraftment.

OPEN IN VIEWER

FIG. 2.

Laboratory and experimental data. (A) May-Grünwald-Giemsa stain of bone marrow aspirate smears (×400). The centrally placed macrophage shows hemophagocytosis of erythroblasts and, possibly, also myeloid cells. Surrounding myelopoiesis presents with an increased proportion of immature forms (left-shift), signs of activation (eg, vacuolated plasma and prominent granulation, with one additional activated macrophage present in lower left corner), and high overall cell turnover, resulting in numerous poorly preserved cell elements. (B) The third-party MSCs were tested regarding their T-cell suppressivity in a 5-day mixed lymphocyte reaction (MLR). T-cell proliferation was assessed based on the thymidine incorporation [counts per minute (CPM)] and compared to unstimulated resting lymphocytes (RLC). (Ci) Serum cytokine profile 10 min before MSC infusion was assessed by multiplex analysis and compared to data from healthy donors (HD) (n=17–108). (Cii) Subsequent changes in these levels were documented at the time points indicated. (D) Cells from bone marrow aspirates were analyzed by multicolor flow cytometry (FACS, LSRII BD Biosciences) at 2 different time points after MSC infusion, on days +6 and +10, at which time the patient clinically presented with low and higher hemophagocytic lymphohistiocytosis (HLH) activity, respectively. The representative FACS data show the gating strategies for T-cells, NK-cells, monocytes, and macrophages. Frequencies and expression of surface molecules and intracellular cytokines were analyzed subsequently and presented in heat maps. (E) Monocytes and MSCs were cocultured for 5 days in independent experiments (n=7), and levels of surface markers (CD206/HLA-DR) and cytokines (IL-10) indicative of an alternative polarization were evaluated by FACS. Bars indicate the standard error of the mean. *P<0.05. Color images available online at www.liebertpub.com/scd

The defective granule-dependent cytotoxicity (and pathogen clearance) in FHL promotes a persistent immunostimulation characterized by a profound hypercytokinemia responsible for most of the clinical manifestations [15]. In the patient's serum, several proinflammatory cytokines and chemokines, the majority of which are already associated with the pathophysiology of FHL [16,17], were elevated before MSC infusion (Figs. 1 and 2C). Upon treatment with MSCs, most of the parameters tested declined within 48 h. Notably, a profound decrease in IL-15 was observed (Fig. 2C). Following an initial peak 3 h after MSC administration, IL-15 concentrations dropped continuously—to almost undetectable levels 72 h later (Fig. 2C). IL-15 is a potent growth factor for T-cells and NK-cells produced mainly by activated mononuclear phagocytes, and associated with inflammatory conditions [18]. This observation warrants further investigations on the role of IL-15 in HLH, as well as studies focused on how MSCs may modulate IL-15. Furthermore, IL-10 was elevated in the patient, consistent with animal models of FLH [19]. Infusion of MSCs further increased serum IL-10. The abundance of anti-inflammatory IL-10 in FHL is considered to be an inherent compensatory mechanism of systemic hyperinflammation. Moreover, it is well established that MSCs promote IL-10 production in both lymphoid and myeloid cells (Fig. 2E) [1,20]. In contrast to the anticipated reduction in the proinflammatory mediators tumor necrosis factor (TNF)-α [20], IL-15, and IL-17, we noticed a transient increase in IL-6 and IL-8 after MSC treatment (Figs. 1 and 2C). MSCs produce and release both IL-6 and IL-8 under inflammatory conditions [21], through a process known as inflammatory licensing [22]. As a part of the MSC-mediated immunoregulation, the abundant IL-6 secreted by MSCs can interfere with the maturation of DCs and can promote the intrinsic production of PGE₂ [1].

Macrophages in the bone marrow, one of the main sites of disease activity, almost exclusively had an immunoregulatory CD163⁺CD206⁺ M2 phenotype (Fig. 2D) [23]. Nowadays, it is speculated that these regulatory macrophages, which produce high amounts of IL-10, may be a reaction to the systemic inflammation rather than its cause. Moreover, recent publications [15] corroborate our in vitro (Fig. 2E) and ex vivo observations, suggesting that MSCs promote such immunomodulatory macrophages. Bone marrow macrophages retrieved at day +6 (when there was clinically low FHL activity) produced less IFN-γ and TNF-α (Fig. 2E) and had a reduced HLA-DR density (data not shown) than at day +10 (when there was clinically higher FHL activity). Increased induction of suppressive macrophages by MSCs would support the rationale of using MSCs to treat FHL.

Overall, the immunomodulatory therapy with MSCs may be beneficial for certain patients with refractory HLH. If transient effects of MSCs in FHL can be confirmed, this therapy may help to bridge the time until allo-SCT is performed and to reduce the need for cytotoxic chemotherapy. Further studies are necessary to answer these questions.