“A Dangerous Black Box:” Idiopathic Hemophagocytic Lymphohistiocytosis in Adult Patients—A Case Report and Review of the Literature

Hemophagocytic lymphohistiocytosis (HLH) is a rare potentially life-threatening condition characterized by aberrant inflammation that can be related to genetic or sporadic forms. In both forms, triggering factors may be involved. Early detection of the underlying cause is crucial for therapeutic decision, while early intervention might be associated with better outcomes. The largest descriptions in the literature on HLH refer to pediatric cases. Adolescents and adults may also be affected, but there is scarce evidence regarding their diagnosis and management. We describe here the case of a 68-year-old Swiss woman with HLH, in whom an extensive search for underlying causes was performed, but neither trigger nor pathogenic variant was found. An early intervention first with dexamethasone and later with cyclosporine was performed. The patient showed a favorable response and did not require further hospitalization; however, one year after diagnosis, it was not possible to suspend cyclosporine due to recurrence of laboratory inflammation signs by drug tapering. The occurrence of HLH idiopathic forms represents a challenge; failure to identify the underlying triggering cause generates uncertainty, endless diagnostic investigations, and consequently additional delays in the treatment. This manuscript addresses the difficulties on this issue.


Introduction
Hemophagocytic lymphohistiocytosis (HLH) is a rare and potentially life-threatening condition characterized by a cytokine storm that is provoked by aberrant activation of macrophages and T cytotoxic cells [1,2]. Clinically, HLH patients present with fever, splenomegaly, and a sepsis-like syndrome. Te laboratory fndings are hyperferritinemia, bicytopenia or pancytopenia, activation of the coagulation cascade, hypertriglyceridemia, often hepatitis, and multiorgan dysfunction [3][4][5]. Pathophysiology is complex: key factors are the presence of predisposing immunodefciency and signifcant immune activation [6]. Although sepsis can trigger HLH, it is important to identify underlying associated HLH as soon as possible, because immunosuppressive therapy is required to suppress the cytokine storm of HLH, a strategy that is not present systematically in sepsis guidelines [7,8].
Te mortality of HLH is highly dependent on provoking conditions and lies between 20% and 88% [9][10][11]. Tere are two forms of HLH: the primary/genetic one, more frequent in children, is caused by pathogenic variants in genes involved in lymphocytic cytotoxicity and immune regulation. Te secondary form, more frequent in teenagers and adults, is associated with diferent medical conditions, such as infections with Epstein-Barr virus (EBV) or cytomegalovirus (CMV), malignancy, and rheumatologic diseases and more recent cases reported due to SARS-CoV-2 infection or vaccination or after chimeric antigen receptor T-cell (CART) therapy infusions among others [6,[12][13][14][15][16][17]. Furthermore, in a variable number of patients, despite extensive investigations,   Case Reports in Hematology no trigger has been found to explain HLH and they will be considered as idiopathic forms. Te latter represents a particular challenge since physicians often might insist on endless diagnosis searches that could lead to an unacceptable delay in the treatment. HLH is a condition better known in children than in adults [1], although awareness regarding HLH in the adult population is increasing. Te incidence of HLH in the pediatric population is 1 in 300.000 [12,13,18], whereas the real incidence of HLH in the adult population remains unknown; however, in the last 10 years, the number of cases reported in the literature in the adult population has signifcantly increased [4,19,20]. Diagnostic tools such as HLH-2004 [21] are based on pediatric populations, and despite being applied to adult patients, their interpretation must be performed with caution [22]. In contrast, the HLHscore (HScore), which was published in 2014, is based on an adult population [23] and relies on the summation of points assigned for 9 variables (immunosuppression, fever, organomegaly, triglyceride, ferritin, glutamic oxaloacetic transaminase, fbrinogen, cytopenia, and hemophagocytosis features in bone marrow aspirates), which correlate with the probability of having HLH. Te Hscore ranges from 0 to 337, giving a probability of having HLH ranging from <1% to ≥99% [23].
Tere are recommendations for the treatment of HLH in pediatric patients, and for example, HLH-2004 [24], an established treatment protocol for adult HLH patients, is however missing. Tus, treatment of adult HLH patients follows existing recommendations for pediatric patients with a variety of nonstandard adaptations. It is common practice to start the treatment with glucocorticoids and another immunosuppressive drug, such as cyclosporine. Te use of etoposide, as recommended by HLH-2004, despite being frequently used, is still discussed in adult patients. Allogenic stem cell transplantation has rarely been reported in adults with secondary HLH [25].
A literature search was performed in PubMed using MeSH terms (performed on January 23 rd , 2022; HLH adult). Among the 1,192 results, 28 cohorts including more than 10 adult patients published from 2010 were identifed and evaluated (Table 1).

Case Presentation
We report the case of a 68-year-old Swiss woman who presented to a peripheral hospital in Switzerland after consulting her general doctor about loss of weight (8 kg) in the last month, a maculopapular rash, fatigue, loss of appetite, palpitations, headaches, and cough for 2 weeks. Te general doctor assigned her to the peripheral hospital after measuring 39.5°C fever and noticing very high liver parameters in the blood test. She had a history of pericarditis at the age of 65 years, and paroxysmal atrial fbrillation and depression were also noted.
Te family history was negative for hematological diseases.
Te laboratory workup at admission showed normocytic, normochromic nonregenerative moderate anemia, mild thrombocytopenia, high D-dimer, slightly lower fbrinogen, liver insufciency, and high triglycerides. Te ferritin level increased (7,069 μg/L, reference value: 10-120 μg/L). An ultrasound scan of the abdomen showed splenomegaly (12.5 cm). Te SARS-CoV-2 polymerase chain reaction (PCR) test, chest X-ray, lumbar puncture, blood culture, serology for EBV, CMV, human immunodefciency virus (HIV), hepatitis A, B, C, and E, Coxiella, parvovirus B19, and Bartonella, and search for Mycobacterium tuberculosis (QuantiFERON test) were negative. After one week of hospitalization, her clinical condition deteriorated, and she was transferred to our center with the suspicion of HLH. At admission, the patient presented with markedly reduced general conditions; she was febrile (38.6°C), hemodynamically compensated, her BMI was 18.5 kg/m 2 , and the laboratory tests showed a progressive increase in CRP up to 190 mg/l, hypoosmolar hyponatremia, increased liver values, coagulopathy, and pancytopenia (Table 2). An HScore for reactive HLH was calculated, showing 209/337 points with an estimated probability for HLH of 93% (Table 3). Terefore, therapy with dexamethasone 15 mg/day was immediately started. Additional investigations were performed, including transthoracic echocardiography, computed tomography (CT) of the chest, abdomen, and pelvis, and a biopsy of the liver, skin, and bone marrow; however, none of these examinations contributed to the identifcation of an underlying trigger for HLH. Te bone marrow showed a large number of macrophages but did not show any sign of hemophagocytosis. Furthermore, a rheumatologic disease could be excluded based on negative screening for autoantibodies (antinuclear antibodies, Hep-2, SsA, PR3 antineutrophil cytoplasmic antibodies, and myeloperoxidase antineutrophil cytoplasmic antibodies) and missing related symptoms. After 1 week of treatment, slight clinical and laboratory improvements allowed doctors to start tapering dexamethasone, and the patient was discharged home after 4 weeks of hospitalization.
Four days later, she presented to her scheduled control in the outpatient clinic of hematology with worsened fatigue. Te laboratory workup showed an increase of ferritin to 49′036 mcg/L while still being under 5 mg dexamethasone/day. We therefore increased the dexamethasone dose to 16 mg/day and started with cyclosporine, with a trough level goal between 100 and 200 ng/ ml. Because up to that point no cause of HLH could be identifed, the investigation was completed with positron emission tomography-computed tomography (PEC-CT), which showed intensive metabolic activity behind the right patella and in the ascending and sigmoid colon. Gastroscopy and colonoscopy performed afterward excluded gastrointestinal neoplasia and local infammatory diseases. Additionally, leg magnetic resonance imaging (MRI) was performed, which suggested pigmented villonodular synovitis (PVNS), a benign entity unlikely to cause HLH, and thus considered an incidental fnding.
Only after 3 months from the manifestation of symptoms, the patient reported a signifcant improvement in her general condition, including regaining appetite and weight.
Te diagnostic workup is summarized in Table 4. Vitamins and substrates (folic acid, iron, zinc, and vitamin B12) were administrated to sustain hematopoiesis.
Six months after the diagnosis, the patient was completely back to her routine life with no symptoms and completely normalized laboratory fndings (Table 2). Slow tapering to discontinue cyclosporine was performed, but immediately before planned drug withdrawal, laboratory signs of infammation with increased liver values at a 9monthfollow-up were observed, and a low dose of cyclosporine was reinstalled. Currently, one year after the diagnosis, the patient is doing well, and tapering of cyclosporine is still ongoing.

Discussion/Conclusion
Several causes have been described as possible triggers of HLH in adult patients (Table 4). Nevertheless, the case presented here did not show any underlying trigger despite an exhaustive search.
Te most important contribution of bone marrow investigations in HLH is the diagnosis or exclusion of hematological diseases. In the bone marrow cytology and biopsy of the case reported here, there was no evidence of lymphoma or other hematological diseases, which may have triggered HLH, and there was also no evidence of hemophagocytosis. Indeed, hemophagocytosis in the bone marrow is reported in just 25% of HLH cases [26]. It is important to note that fnding a hemophagocytosis phenomenon in the bone marrow is neither specifc nor sensitive for HLH [27]. Furthermore, in our case, a skin biopsy was performed to rule out Sézary syndrome due to the presence of maculopapular exanthema erythematosus of unknown origin. Likewise, due to the persistence of liver test abnormalities, a liver biopsy was also performed. Te results of both biopsies did not show lymphoma or other relevant fndings, and in both samples, no hemophagocytosis phenomenon was observed.

Case Reports in Hematology
Performing PET-CT on HLH patients contributes to the diagnosis of lymphoma and other malignant or infammatory diseases [28][29][30].
After excluding a secondary cause of HLH in our patient and despite the low probability due to age, as well as the lack of positive family history, we considered the possibility of primary/genetic HLH. Tis form is better known in children and often associated with other abnormalities such as immunodefciency or albinism, such as in Griscelli syndrome [31]. Panel sequencing of HLH-associated genes or exome sequencing have made a great deal of contribution to the diagnosis of congenital forms, particularly in children. In fact, with these investigations, molecular explanation could likely be found in 58% of the studied pediatric patients even in the absence of familial syndromes [32]. Te most common pathogenic variants reported in 14% of the cases associated with HLH in adults were identifed in PRF1, STXBP2, and UNC13D [33]. In our patient, the analysis of these genes was negative.
With the frst suspicion of HLH, despite not knowing the triggering cause, an early intervention frst with dexamethasone and later with cyclosporine was performed, as suggested by other authors [1,5]. Te patient initially responded well to glucocorticoid monotherapy, but while tapering at one month (Table 2), she had a relapse. A careful glucocorticoid tapering method combined with another immunosuppressive (cyclosporine) drug was afterward successful in this case.
In the literature review, we found 28 case series or studies published after 2010 based on the adult population [11,26,28,30,. Cohorts were collected from different countries and reported on a number of patients, from 12 [46] to 264 [38]. A meta-analysis [11] analyzing 2,197 patients was also included. We found that the most common trigger reported for secondary HLH in adults was malignancy, both solid cancer and hematological diseases, with a weighted average of 50% (from 0% to 72%). Te second most common trigger was infection with 39% (from 0% to 76%) and autoimmune diseases with 10% (0% to 58.3%). Idiopathic HLH was reported in 24/28 studies, and the weighted average was 9.4% (0% to 50%). Te diferent observed prevalence of triggers may be explained by the diferences in the median age of the studied population. Furthermore, not all centers had the same possibility and easy access to diagnostic tools such as MRI or PET-CT and therefore were not included in publications. Tis diversifcation in patients and used diagnostic methods may partially contribute to the variability of the prevalence of idiopathic HLH.
Regarding the mortality, some studies reported just the 30-day mortality, others the mortality during the hospitalization or for the whole follow-up. For this analysis, we focused on the 30-day mortality, which was 38%, and on the overall mortality, which was 53%. In addition, mortality may be infuenced by diferent factors, such as the prevalence of diferent underlying diseases, comorbidities, and age of the observed population, as well as the experience and resources of the center in diagnosing and treating HLH rapidly.
HLH is a life-threatening condition and requires rapid diagnosis and therapy. Unknown origin of HLH may be frustrating for both physicians and patients. According to our experience, it is very important to perform a complete diagnostic test, including tissue biopsy, diferent imaging studies, and genetic analysis whenever possible. All possible triggers should be searched because specifc treatment of the underlying "trigger" disease may play a key role in management [5]. Every diagnostic and therapeutic step must be explained clearly to the patient to keep high compliance, which is essential in this process that may last several months in some cases. Our patient was very consistent with the prescribed medications and the appointments in our outpatient clinic, and she also had great support from her husband. Tey understood how dangerous her condition was; nowadays, they are happy with the outcome and the treatment received, and the patient is still under follow-up. A close follow-up is recommended after the frst episode of idiopathic HLH to allow early diagnosis of an eventual relapse.
With currently available diagnostic methods, idiopathic forms of HLH still exist. Considering the high reported rate of mortality in many series, it is important to emphasize that in seriously afected patients, the lack of a trigger should not delay the initiation of immunosuppressive treatment, which should be started as soon as possible and should be complemented by the search for underlying causes.
We can conclude that despite the recent contributions of some publications, there is clearly a defcit in the understanding of HLH in adults. Due to the rarity of its presentation, only joint eforts of many centers investigating this disease in a prospective way will better identify its epidemiological, diagnostic, and therapeutic aspects, allowing for tailored management of HLH adults in the future.

Data Availability
Data analyzed in this study were reanalysis of the existing data, which are openly available at locations cited in the reference section.

Ethical Approval
Ethical approval is not required for this study.

Consent
Written informed consent was obtained from the patient for publication of this case report and any accompanying images.