Real-world experience with targeted therapy in patients with histiocytic neoplasms in the Netherlands and in Belgium

Abstract

Histiocytic disorders are rare hematologic neoplasms characterized by a notable dependence on mitogen-activated protein kinase signaling. Targeted therapy is an emerging treatment option, yet the number of reported patients remains limited. Here, we describe 40 patients with histiocytic neoplasms who were treated with targeted therapy in 7 tertiary referral hospitals from the Netherlands and Belgium. The cohort comprised of 6 (15%) children and 34 (85%) adults with diverse histiocytoses, including Langerhans cell histiocytosis (LCH; n = 12), Erdheim-Chester disease (n = 14), central nervous system xanthogranuloma (n = 2), Rosai-Dorfman disease (n = 3), histiocytic sarcoma (n = 2), ALK-positive histiocytosis (n = 1), and mixed/unclassifiable histiocytosis (n = 6). Five patients were included in a clinical trial; 35 (88%) received BRAF/MEK inhibitors outside of trials. Among these 35 patients with available follow-up data, median time on targeted treatment was 1.9 years (range, 0.04-5.8 years). Complete or partial responses were observed in 25 of 27 (93%) patients treated for multisystemic and/or solid lesions and 2 of 8 (25%) patients treated for neurodegenerative LCH. Responses were generally durable, although 10 patients lost response after dose reduction or therapy interruption. Responses were recaptured in 9 of 10 cases. Two patients developed new or progressive neurodegenerative lesions: 1 during and 1 after vemurafenib therapy. At last follow-up, 8 adults had stopped targeted therapy because of toxicity. This study corroborates the favorable outcomes of BRAF/MEK inhibition in patients with histiocytosis described previously. However, it also highlights limitations and calls for prospective studies.

Graphical abstract

Figure 1.

Patient and treatment characteristics. Pie charts depicting proportions of patients by (A) age category (with age measured in years at start of targeted therapy), (B) histiocytosis subtype, (C) inhibitor subtype (at start of targeted therapy), or (D) clinical trial enrollment.

Figure 2.

Swimmer plot depicting targeted therapy outcomes. Patients are grouped by disease subtype. For each patient, targeted therapy is represented by horizontal bars, with the size of these bars proportional to the length of treatment, and the color representing the type of treatment. Targeted therapy was initiated at time point zero. To the left of the horizontal bars, best response on targeted treatment is summarized for each patient; when applicable, separately for multisystemic/solid lesions and neurodegenerative lesions. Within the bars, outcomes of radiologic response assessments are shown as colored rectangles. In the absence of radiologic response assessments, clinical responses are visualized as colored ovals. In general, the first assessment establishing a particular response (eg, a PR) is depicted. AXG, adult xanthogranuloma; BRAFi, BRAF inhibitor; H-UC, unclassifiable histiocytosis; HSCT, hematopoietic stem cell transplantation; JXG, juvenile xanthogranuloma; MS, multisystem; PEG-IFN-α, pegylated interferon-α.

Figure 3.

Response and rare disease progression on targeted therapy. (A) Photographs showing response of skin lesions in case 30 with RDD/ECD to conventional therapies and targeted treatment with cobimetinib. (B) Axial PET-computed tomography (CT) images of case 17 with ECD showing increased fluorodeoxyglucose (FDG) uptake of a left acetabular lesion (indicated by a white arrow) during intermittent dosing of dabrafenib/trametinib. After returning to continuous therapy, PET-CT showed decreased FDG uptake in the acetabular lesion, consistent with a recaptured PR. (C) Axial T1-weighted gadolinium-enhanced MRI images of case 36 with CNS-RDD showing an initial PR of contrast-enhanced brain lesions to treatment with cobimetinib, and subsequent progression of brain stem lesions after dose reduction to 20 mg/d. After return to full dose (60 mg/d), lesions decreased again, indicating a recaptured response. (D) Axial CT and PET images of case 16 with ECD showing progression of an FDG-avid lung nodule during treatment with dabrafenib, and partial resolution of this nodule after switch to treatment with trametinib. 2-CDA, cladribine; DEX, dexamethasone; BRAFi, BRAF inhibitor; MEKi, MEK inhibitor.

Figure 4.

Disease progression after targeted therapy interruption. (A) Sagittal and axial MRI images of case 28, with CNS-AXG showing response of brain lesions to conventional therapies and targeted treatment with vemurafenib. After an initial PR to vemurafenib, targeted treatment was interrupted because of discontinued insurance coverage. Hereafter, brain lesions slowly progressed. Vemurafenib was restarted 3.6 years after interruption and quickly resulted in a decrease of brain lesions, with almost no lesions discernible at last follow-up. (B) Sagittal MRI images of case 27, with CNS-JXG showing relapse of a contrast-enhanced intramedullary spinal cord tumor after targeted therapy interruption. In this young child, treatment with dabrafenib and trametinib was interrupted after 1.6 years of therapy, while in complete radiologic remission. As an exit strategy, trametinib was stopped first, followed by dabrafenib 2 weeks later. MRI after stop of trametinib revealed persistent radiologic CR. However, MRI 2 weeks after stop of dabrafenib revealed recurrence of contrast-enhancement, consistent with a relapse of disease. MRI 2 weeks after restart of dabrafenib and trametinib showed that the radiologic response was recaptured. JXG, juvenile XG; MPS, methylprednisolone.

Figure 5.

Response and progression of ND lesions. (A) Axial MRI images of case 6, with isolated ND-LCH showing significant reduction in T2-hyperintense lesions after treatment with dabrafenib and trametinib. Note that the remaining hyperintense lesion in the right cerebellum at last follow-up (right column; middle image) is related to the biopsy that was taken from this anatomical site. (B) Axial MRI images of case 33 showing the development of T2- and FLAIR-hyperintense lesions in the pons and cerebellum while receiving treatment with vemurafenib. The patient also developed clinical symptoms of neurodegeneration and died 7.5 months after the MRI scan depicting ND lesions was made. ND, neurodegenerative; FLAIR, fluid-attenuated inversion recovery.

Figure 6.

Overall survival and causes of death. Kaplan–Meier curves depicting overall survival for children and adults treated with targeted therapy outside clinical trials. Five of the 35 patients died, including 3 from disease-related causes; “1” is case 29, “2” is case 26, “3” is case 32, “4” is case 33, and “5” is case 9.