HLH is a rare but devastating clinical entity and has been associated with several rheumatologic disorders including adult Still’s disease, sarcoidosis, systemic sclerosis and Sjogren’s syndrome . An extensive literature review was completed for articles published up to September 2017 based on a bibliographic search in the PubMed database using the keywords “Hemophagocytic lymphohistiocytosis” and “Systemic lupus erythematosus” with inclusion criteria of articles focusing on cutaneous manifestation. Case reports involving new diagnosis of SLE as the onset of HLH is limited [8,9,10,11,12]. In the cases presented here, case 1 and 2’s clinical onset of HLH did coincide with the new onset of SLE and does fulfill the SLICC criteria for SLE. Amongst all cases of SLE and HLH, including rheumatologic flares, the incidence of HLH is approximately 0.9–4.6% . As has been described by several groups, the difficulty in making the diagnosis of HLH in new onset SLE arise from the overlap of many symptoms, making the use of the HLH-2004 criteria critical for acute and timely diagnosis. Several other parameters have been used to tease out the elusive diagnosis, however with mixed consensus in the literature. Hyperferritinemia has been cited as the best parameter to distinguish between active SLE flare and HLH-associated SLE with a sensitivity and specificity of nearly 100% . However, hyperferritinemia in the HLH-94 study indicated that in the pediatric population, ferritin level > 500 mcg/L was 100% sensitive for HLH but less specific; whereas ferritin > 10,000 mcg/L was 90% sensitive and 96% specific for HLH . In the adult population, the correlation between hyperferritinemia and HLH is less clear, particularly with the presence of an overlapping autoimmune condition such as SLE. One study evaluated HLH-associated SLE with solely active SLE and concluded a high ferritin level often points towards the former .
More recently, a group from France has published the “HScore”, a well-validated scoring system for the diagnosis of HLH . The scoring system incorporates parameters including organomegaly, ferritin, ALT, degree of cytopenia, fibrinogen, fever, and hemophagocytosis with a HScore > 250 conferring 99% probability of HLH and a score < 90 at < 1% probability. Interestingly, for case 1 the HScore was calculated as 245 (conferring 99.1% probability of HLH) and case 2 as 201 (conferring 88.8% probability for HLH). In the future, the HScore may be a useful clinical parameter to help tease out active SLE from HLH-associated SLE.
The two cases here highlight an extremely rare entity within HLH-associated SLE with severe cutaneous manifestation at the time of diagnosis. A case report from Japan also observed this entity with erythematous plaques as the initial manifestation of HLH in a patient with newly diagnosed lupus . Furthermore, HLH can present with skin manifestation but (in) the absence of antibodies can eliminate an associated autoimmune process.
In patients with known SLE, the parameters to gauge flare from HLH is even more difficult. One study assessed patients with known autoimmune conditions (including SLE) and performed skin biopsies on three patients during acute flare, of which all were found to have hemophagocytosis . While this may be an important additional clue to help tease apart the two conditions, hemophagocytosis is a non-specific finding and has been associated with self-limiting bone marrow infections in the context of activated macrophages. A laboratory study assessed hemophagocytosis by comparing bone marrow aspirates in patients with known HLH compared to random control bone marrows . The sensitivity of hemophagocytosis in HLH was 83% with a specificity of only 60%, suggesting rare hemophagocytes may be seen in normal bone marrow. The authors suggest a rise of hemophagocytosis count threshold to increase to 0.05–0.13% to account for rare ‘normal’ markers would increase specificity to 100%, which may also aid in teasing out HLH-associated SLE from SLE flare.
Amongst all currently available immunologic studies available, sIL-2R appears to correlate best with disease activity. One study assessed patients with lymphoma-associated HLH compared to non-lymphoma cases and found that the former had a much higher sIL-2R to ferritin ratio (8.56 vs. 0.66) . These ratios have not been assessed in autoimmune associated HLH, specifically SLE, and may provide a promising avenue to better discern this diagnosis.
Classically, treatment of primary HLH is directed towards use of the HLH-2004 protocol including etoposide, dexamethasone, cyclosporine, consideration of intrathecal methotrexate and finally with hematopoietic stem cell transplantation . Treatment of secondary HLH is less clear, but ultimately relies on treatment of the primary auto-immune disease. In the two cases presented, SLE was the process associated with HLH and corticosteroids were rapidly initiated, in keeping with previously described cases . Because of the high mortality risk of HLH, the absence of rapid improvement of symptoms, even in the context of non-severe manifestations of SLE, requires aggressive immunosuppression with cyclophosphamide and frequently the addition of biologics such as Anakinra (Il1 inhibitor), infliximab (TNF inhibitor)  and alemtuzumab (CD52 inhibitor) [6, 21, 22].
Based on the cases and literature review presented here, we suggest considering HLH in the differential diagnosis with adult patients presenting with new onset, severe cutaneous manifestation of lupus and to consider appropriate investigations and prompt treatment.