BRAF-V600E–mutated Rosai-Dorfman-Destombes disease and Langerhans cell histiocytosis with response to BRAF inhibitor

The histiocytoses are rare neoplasms/disorders caused by accumulation of abnormal cells derived from macrophages, dendritic cells, or monocytes.¹  Diagnosis is based on the combination of morphologic and immunophenotypic findings, molecular alterations, clinical presentation, and imaging. Langerhans cell histiocytosis (LCH) and Rosai-Dorfman-Destombes disease (RDD) exist in the L-group and R-group, respectively, based on recent reclassification of histocytoses.²  Although there are a handful of mixed RDD/LCH reports (unassociated with other neoplasia),^3-11  the concept of “mixed histiocytosis” has varied definitions, including lesions either simultaneously occurring in the same location with distinct microenvironments vs those lesions in the same patient at different time points.

LCH is a clonal proliferation of aberrant myeloid-derived dendritic cells (CD1a/S100/Langerin-positive).^1,12  Cases with BRAF-V600E mutation have overexpression of the VE1 protein immunohistochemical stain with good molecular correlation.¹³  The BRAF-V600E mutation has been associated with both increased risk of relapse^14,15  and more permanent, long-term consequences including central nervous system (CNS) neurodegenerative disease.¹⁵  Quantitative assessment of blood for BRAF-V600E mutation offers a noninvasive, reliable modality for monitoring patients during treatment.¹⁶ 

RDD, histologically characterized by large histiocytes with ample pale cytoplasm (S100/fascin-positive),¹  often with emperipolesis, presents not only with lymph node (LN) enlargement, usually of the cervical chain, but also extranodal disease, including the retro-orbital tissue and CNS lesions.^17,18  RDD is often self-limited, but the number of systems involved determines prognosis.^2,17  Although there is no standard therapy, recent RDD consensus treatment guidelines outline when to consider MAPK therapy.¹⁷  Two recent reports of localized RDD have demonstrated BRAF mutations.^19,20 

We report a child with simultaneous mixed systemic RDD and LCH, both with BRAF-V600E mutation, and subsequent clinical and radiographic response after treatment with the BRAF-inhibitor therapy, dabrafenib.

Histopathologic features were assessed using hematoxylin-and-eosin–stained sections (3 µm) from formalin-fixed, paraffin-embedded tissue and immunohistochemistry was performed on a Ventana platform using commercially available antibodies. Immunohistochemistry analysis for BRAF-V600E mutation was performed on formalin-fixed, paraffin-embedded tissue sections using the anti-BRAF V600E (VE1) mouse monoclonal antibody and OptiView DAB detection system (Ventana Medical Systems, Tuscon, AZ). Concurrent positive and negative controls showed the expected results. Staining interpretation is based on qualitative interpretation, as previously described.¹³ 

Assay of BRAF p.V600E mutation on peripheral blood mononuclear cells (PBMCs) was performed (Texas Children’s Hospital, Baylor Hospital, Houston, TX) using a quantitative analysis platform previously described, which can detect mutant cells at 0.01% or greater frequency.¹⁴ 

A 6-year-old previously healthy child had several months of neck pain and cervical lymphadenopathy without systemic symptoms or laboratory abnormalities and was diagnosed with both RDD and LCH by histologic assessment of an LN biopsy. Neck magnetic resonance imaging (MRI) showed C4 vertebra plana and bilateral cervical lymphadenopathy. A positron emission tomography scan showed multiple fluorodeoxyglucose (FDG)-avid LNs in the neck, axilla, subcarinal, and supraclavicular areas and bone lesions in the maxillary sinus, C4 vertebrae, and right iliac wing (supplemental Figure 1B). Brain MRI showed T2 abnormalities in the bilateral cerebellar hemispheres, basal ganglia, and corpus callosum with soft tissue infiltration of the maxillary sinus and no pituitary involvement.

The diagnostic cervical LN core biopsy showed sinus expansion with RDD histiocytes (S100/fascin-positive), frequent emperipolesis, and inconspicuously intermixed sinus LCH-like histiocytes (S100/CD1a/focal Langerin) that were best highlighted on CD1a with plump clusters of sinus-based LCH cells with typical nuclear features (Figure 1). No distinctive RDD histiocyte showed expression of CD1a or Langerin. Strong BRAF-V600E (VE1) antibody staining (3+) was seen in both histiocyte types (Figure 1). Circulating BRAF-V600E mutation was detected in PBMCs at diagnosis and throughout his course (0.07% to 0.02%), without evidence of molecular remission (supplemental Figure 1A).

Due to cerebellar involvement, consistent with LCH, along with tissue diagnosis, the patient was started on cytarabine (170 mg/m² × 5 days every month) based on published experience.²¹  Disease reevaluations after cycles 2 and 4 of cytarabine showed mixed response, with complete resolution of bone disease typical for LCH but no interval change in cervical LN disease, and increased FDG uptake in other LN disease typical of RDD. Brain MRI showed no significant changes from baseline. Cervical LN rebiopsy following the fourth cycle of cytarabine showed a similar mixed population of residual RDD/LCH with strong VE1 expression in both. Therapy was changed to BRAF-inhibitor dabrafenib 5.25 mg/kg per day for persistent nodal disease and continued circulating PBMCs with BRAF-V600E mutation (supplemental Figure 1A). The patient has tolerated dabrafenib well besides mild arthritis, successfully treated with naproxen. Following 3 months of BRAF-inhibitor therapy, imaging showed decreased size and FDG uptake of cervical LN disease with resolution of additional FDG-avid LN disease (supplemental Figure 1B). Brain MRI showed improved signal abnormalities. However, after 13 months on monotherapy with dabrafenib, there was a slight increase in frequency of circulating BRAF-V600E PBMCs (supplemental Figure 1A) and minimal increased FDG uptake in cervical disease.

This case illustrates 3 concepts with respect to RDD, mixed histiocytosis, and treatment with BRAF-V600E inhibitor therapy.

First, this is the third reported case of RDD harboring an identifiable BRAF mutation,^19,20  and first reported example of BRAF-V600E mutation identified simultaneously in RDD and LCH, as a systemic mixed histiocytosis in a pediatric patient. Although reports of MAPK pathway mutations in RDD have been described with growing frequency,^22-25  the BRAF-V600E mutation has only been reported in 1 localized RDD, diagnosed initially by BRAF VE1 immunostain, similar to our case. In that case, molecular confirmation was made only with multiplex picodroplet digital polymerase chain reaction, as it was originally negative by pyrosequencing.¹⁹  In our case, molecular confirmation of BRAF-V600E was made on circulating PBMCs with clinical response after dabrafenib therapy.

Second, the current histiocytic classification system identifies the R-group and the L-group as separate entities based on pathology and molecular and clinical disease involvement; however, there is little mention of mixed histiocytosis between these 2 groups. The handful of reported instances of RDD and LCH both in the same patient lack a formal definition of what constitutes “mixed histiocytosis.” These mixed LCH/RDD histiocytosis cases suggest a shared pathogenesis. Although previous descriptions of a “transitional cell” were made in such cases (ie, cytomorphology of RDD and expression of CD1a and Langerin),^3,10  our case had no such cell. Rather, the close intermingling of LCH/RDD cells in the sinuses, decreased Langerin expression in LCH cells, and dual expression of BRAF VE1 in both histiocytes may suggest a common BRAF-V600E precursor that can differentiate along 2 different pathways as a result of as-yet-undefined influences. In adults, various presentations of mixed histiocytosis with LCH/Erdheim-Chester disease (ECD) sharing a BRAF-V600E mutation have been reported and is 1 reason for reclassifying them as L-group lesions.^2,26  However, in rare adult cases of mixed R- and L-group lesions (ie, isolated RDD with ECD²⁷  or LCH-ECD-RDD with myeloid fibrosis²⁸ ), none have demonstrated a BRAF-V600E mutation and only a few demonstrate other MAPK mutations.²⁷  In reported mixed RDD/LCH cases (Table 1), none have documented BRAF mutations.

Third, treatment with dabrafenib resulted in sustained clinical and radiographic improvement. However, the slight increase of BRAF-V600E in PBMCs following 13 months of dabrafenib therapy, along with increased positron emission tomography uptake in the cervical nodes, raises concern for relapse. It is known that BRAF inhibition in histiocytic diseases is not curative in most cases (ie, not cytotoxic) but does achieve rapid and sustained clinical remission, as demonstrated in this case.^16,29-31  This highlights the need for prospective studies to evaluate BRAF-inhibition dosing regimens, optimal treatment duration, and monotherapy vs combination therapy.¹⁶  Although the efficacy of BRAF and MEK inhibition has been well documented in the adult histiocytoses,^32,33  there are fewer reports of efficacy in the pediatric population, mostly limited to case reports/series.^16,29-31  In this case, given the mild increase in the peripheral BRAF-V600E and FDG uptake, MEK inhibition with trametinib (0.025 mg/kg by mouth) was also recently added, with disease reevaluation ongoing. Currently, the patient has been on both dabrafenib and trametinib for 6 weeks as of this writing with follow-up scan after 12 weeks of dual therapy.

The authors thank Lori Schmitt at UPMC Children’s Hospital of Pittsburgh for assistance in preparation of antibody table and technical expertise in immunohistochemistry. The authors also acknowledge the UPMC Children’s Hospital of Pittsburgh Radiology Department for their support, and specifically Judy Squires for assistance in reviewing the imaging for this case.

This work was supported by the National Institutes of Health, National Center for Advancing Translational Sciences through grant number UL1TR001857.

The work was performed at UPMC Children’s Hospital of Pittsburgh. This work was also supported by the University of Pittsburgh Department of Pathology.

Contribution: J.P., A.C., and S.M. were responsible for study concept, design, and analysis; and all authors participated in manuscript preparation, editing, and review.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Jennifer Picarsic, UPMC Children’s Hospital of Pittsburgh, One Children’s Hospital Dr, 4401 Penn Ave, Pittsburgh, PA 15224; e-mail: jenpicarsic@gmail.com.