• anti-tnf
• cardiovascular disease
• outcomes research

Erdheim-Chester disease (ECD) is a rare non-Langerhans cell histiocytosis characterised by long bones and various other organs’ involvements.1 Tissues are infiltrated by CD68+ CD1a− foamy histiocytes.2 Therapeutic options are pegylated interferon-α,3 cladribine,4 mTOR inhibitors5 and anakinra in mild forms of ECD.6 7 Conversely, targeted BRAF or MEK inhibitor therapies are used for refractory or life-threatening ECD manifestations.8 9 All these therapies have frequent side effects. Interferon-α leads to fatigue, cytopenias and autoimmune diseases. Anakinra needs subcutaneous daily injections, with frequent local intolerance. Targeted therapies cause skin cancers, QT allongement, rhabdomyolysis and haemorrhage,10 and we lack long-term safety data with these drugs. Moreover, targeted therapies are not available in all countries. Thus, there is an unmet need for improving treatments in many patients with ECD, particularly in countries where access to targeted therapies is complicated or impossible.

Although our comprehension of this disease has recently moved from inflammatory to clonal myeloproliferative disease,11 the accumulation of pathological histiocytes leads to an increase of several cytokines in blood and affected tissues. Cytokine/chemokine network has been described in ECD lesions, and most of these factors are regulated by tumour necrosis factor alpha (TNF-α), which is also expressed in ECD lesions and serous fluids like pericardial effusion12–14 and has been implicated in ECD pathogenesis in in vitro models.15 Two ECD patients with cardiovascular involvement were successfully treated with infliximab, a TNF-α monoclonal chimeric antibody, achieving an improvement of symptoms, cardiac involvement and function.14 We aim to determine the efficacy of infliximab in a larger series of patients with ECD.

We retrospectively reviewed the medical records of patients with ECD who received at least one infusion of infliximab between 2011 and 2016 in our tertiary care centre. ECD diagnosis was based on clinical and radiological presentation consistent with ECD and histological confirmation as described elsewhere.1 All patients underwent BRAF mutation detection on codon V600 by pyrosequencing as previously described,16 and whenever negative were then analysed by multiplex picodroplet digital PCR as previously described.17 The eligible patients received infliximab 5 mg/kg D0–D15 then every 4 weeks. After 6 months, infliximab was administered every 6–8 weeks. All patients (except those who had a contraindication to this treatment) also received weekly oral methotrexate at the dosage of 10 mg, to prevent immunisation against infliximab chimeric antibodies.

Since ECD is a multisystemic heterogeneous disease, and due to the fact that response assessment was not uniformly performed, we used several criteria for assessing the efficacy: blood C reactive protein (CRP) levels, metabolic responses and radiographic evolution. The metabolic response was classified as complete metabolic remission (CMR), partial metabolic remission (PMR), stable disease or progression. The PERCIST criteria were used to evaluate this response and is extensively described elsewhere.8 All the ¹⁸FDG PET-CT scans were centrally reviewed. Radiographic evolution was classified in progression, stable or regression in successive assessments. All adverse events were noted. The infliximab treatment was continued until severe adverse event, death or disease progression occurred.

Qualitative data were described with numbers and percentages and quantitative values with the medians and ranges. Quantitative values (standard uptake value (SUV) and CRP levels) were compared with the non-parametric Wilcoxon rank-sum test for paired values. All tests were two-sided, and a P value 0.05 was considered to be statistically significant. Statistical analyses were performed using GraphPad Prism V.6.0 software (GraphPad Software, San Diego, California, USA).

Sixteen patients (11 men, 5 women, median age at ECD diagnosis 67; range 16–72) received at least one infliximab infusion. The patients previously received interferon-α (n=11), anakinra (n=3) and/or cladribine (n=2). Seven patients (64%) harboured the BRAF ^V600E mutation (BRAF status was not determined in five patients). All patients had multisystemic involvement with aortic (n=12), retroperitoneal (n=10), cardiac (n=7), central nervous system (CNS) (n=5), skin (n=4) and retro-orbital (n=3) infiltrations. Two patients died before treatment efficacy was assessed: one died rapidly after treatment initiation due to a progression of multisystemic involvement, and another one died from stroke complications 6 months after infliximab was started. One patient prematurely stopped the treatment for personal convenience, and another one was lost to follow-up. Thus the treatment efficacy was evaluated in 12 patients (table 1). The median duration of infliximab treatment was 20 months (11–60). Five patients (42%) achieved PMR, three (25%) had stable disease and four (33%) experienced disease progression under treatment. Median SUVmax in target lesions was not significantly improved between baseline and last evaluation (figure 1A). The blood CRP level was not different between baseline and last visit (median CRP at baseline 17 mg/L; median CRP at last infusion 15 mg/L, P=0.85) (figure 1B). Cardiac involvement, which was observed in 5/12 patients, was stable in 3 patients, worsened in 1 and improved in 1 in radiographic assessments. CNS involvement, which was present in four patients, was stable in two (pseudo-meningioma) and improved in two (cerebellar infiltration). Retro-orbital infiltration was stable in two and worsened in one. Retroperitoneal infiltration was stable in nine and worsened in one. Sinus involvement improved in one patient. Eight patients experienced side effects during the treatment: five had infectious events, one myelodysplasia and one hypogammaglobulinaemia. The treatment was discontinued in 10 patients: 1 for death (ischaemic stroke), 2 for serious adverse events (haematological malignancy and severe infection, the latter had also a progressive disease) and 4 for progression of ECD. Additionally, two patients who had stable disease stopped the treatment for personal reason (they had mild or no symptoms and preferred quitting treatment). One patient had severe CNS involvement, and despite a mild improvement observed under infliximab, she was prescribed vemurafenib after 11 months of infliximab because of the presence of BRAF ^V600E mutation.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Outcomes of patients with Erdheim-Chester disease under infliximab therapy. Standard uptake values (¹⁸FDG PET scans) of target lesions at baseline and end of follow-up are shown in (A), and blood C reactive protein (CRP) levels in (B). (C) The improvement of bone and central nervous system hypermetabolisms in patient #9. (D) The worsening of pleural thickening and effusion under infliximab in patient #8.

Our study demonstrates that infliximab has a variable efficacy in ECD. Metabolic response was achieved in 42% of patients. Cardiac efficacy was variable. CNS involvement improved in two patients. Overall, the toxicity profile was overall favourable with only two patients in whom the treatment was discontinued due to adverse events.

We now have increasing evidence that targeting therapies (BRAF and/or MEK inhibitors) leads to dramatic improvement in BRAF mutated as well as patients with wild-type ECD.8 However, the tolerance profile of such drugs is questionable and they should be kept for patients with refractory disease or life-threatening manifestations. Moreover, the majority of patients relapse after BRAF inhibitors interruption.10 Infliximab could be used in mild forms of ECD or in second line after targeted therapies interruption.

The limitations of our studies are due to the retrospective design and heterogeneity in disease assessment. The ¹⁸FDG PET scans were centrally reviewed and objective criteria were used to determine the overall metabolic response, but the metabolic assessment has some limitations for evaluating anticancer response.

In conclusion, infliximab has a moderate although variable efficacy in ECD. Our study suggests that infliximab has little efficacy in ECD and could be considered as a treatment option in patients with mild disease, intolerance to other drugs or after targeted therapies interruption.