Clinicogenomic associations in childhood Langerhans cell histiocytosis: an international cohort study

Abstract

Langerhans cell histiocytosis (LCH) is a rare neoplastic disorder caused by somatic genetic alterations in hematopoietic precursor cells differentiating into CD1a+/CD207+ histiocytes. LCH clinical manifestation is highly heterogeneous. BRAF and MAP2K1 mutations account for ∼80% of genetic driver alterations in neoplastic LCH cells. However, their clinical associations remain incompletely understood. Here, we present an international clinicogenomic study of childhood LCH, investigating 377 patients genotyped for at least BRAFV600E. MAPK pathway gene alterations were detected in 300 (79.6%) patients, including 191 (50.7%) with BRAFV600E, 54 with MAP2K1 mutations, 39 with BRAF exon 12 mutations, 13 with rare BRAF alterations, and 3 with ARAF or KRAS mutations. Our results confirm that BRAFV600E associates with lower age at diagnosis and higher prevalence of multisystem LCH, high-risk disease, and skin involvement. Furthermore, BRAFV600E appeared to correlate with a higher prevalence of central nervous system (CNS)-risk bone lesions. In contrast, MAP2K1 mutations associated with a higher prevalence of single-system (SS)-bone LCH, and BRAF exon 12 deletions seemed to correlate with more lung involvement. Although BRAFV600E correlated with reduced event-free survival in the overall cohort, neither BRAF nor MAP2K1 mutations associated with event-free survival when patients were stratified by disease extent. Thus, the correlation of BRAFV600E with inferior clinical outcome is (primarily) driven by its association with disease extents known for high rates of progression or relapse, including multisystem LCH. These findings advance our understanding of factors underlying the remarkable clinical heterogeneity of LCH but also question the independent prognostic value of lesional BRAFV600E status.

Graphical abstract

Figure 1.

Clinical features at LCH diagnosis according to BRAF^V600Estatus. (A) Prevalence of BRAF^V600E in patients with specific clinical characteristics at LCH diagnosis. (B) Prevalence of BRAF^V600E in patients with specific types of bone involvement at diagnosis. This figure depicts all patients with osseous lesions, irrespective of single-systemic or multisystemic disease extent. Bones are grouped according to the classification used by the National Cancer Institute. Upper extremity: humerus, radius, ulna, carpals, metacarpals, and phalanges. Shoulder girdle: clavicle and scapula. Pelvic girdle: coxal, innominate, and hip bones (including ilium, ischium, acetabulum, and pubis). Lower extremity: femur, tibia, fibula, patella, tarsals, metatarsals, and phalanges. Spinal column: cervical, thoracic and lumbar vertebrae, sacrum, and coccyx. Thoracic cage: ribs and sternum. CNS-risk lesions are bone lesions affecting the orbital, temporal/mastoid, sphenoidal, zygomatic, or ethmoidal bones, the maxilla, paranasal sinuses, or anterior or middle cranial fossa, according to LCH Study Group definitions.^,, , (C) Prevalence of BRAF^V600E in patients with specific presentations of SS-skin LCH at diagnosis. Numbers of patients are provided in Table 1 and supplemental Table 3. Dashed lines indicate the prevalence of BRAF^V600E in all cases (51%). Statistical tests with P < .05 are shown. ∗P < .05, ∗∗P < .00125. MS, multisystem; SS, single-system; RO, risk organ; UFB, unifocal bone; MFB, multifocal bone; CNS, central nervous system.

Figure 2.

Clinical features at LCH diagnosis of children with BRAF^V600E, BRAF exon 12, or MAP2K1 mutations. (A) Pie chart showing the mutational status of the 377 patients from our cohort. (B) Dot plot showing age at diagnosis of patients with BRAF^V600E, BRAF exon 12, or MAP2K1 mutations. Error bars depict medians with interquartile ranges. (C-E) Bar charts depicting the percentage of patients with BRAF^V600E, BRAF exon 12, or MAP2K1 mutations having specific disease extents at LCH diagnosis. Statistical comparisons were performed for SS-bone disease in panel D and MS disease in panel E. (F-H) Bar charts depicting the percentage of patients with BRAF^V600E, BRAF exon 12, or MAP2K1 mutations having specific disease sites at LCH diagnosis. Statistical tests with P < .05 are depicted. Numbers of patients are provided in Tables 1 and 2. ∗P < .05, ∗∗P < .00125. WT, wild-type.

Figure 3.

Clinical outcome of patients with pediatric LCH according to mutational status and disease extent. Kaplan-Meier curves showing EFS according to lesional BRAF^V600E status for all 373 patients (A), 339 patients with low-risk LCH (B), 285 patients with SS LCH (C), or 183 patients with SS-UFB LCH (D) at diagnosis. Patients with low-risk LCH comprise all patients except those with high-risk (MS-RO⁺) disease. Kaplan-Meier curves showing EFS of patients with BRAF^V600E, BRAF exon 12, or MAP2K1 mutations (E-F). Curves are shown for all 282 patients (E) and for 141 patients with SS-UFB LCH (F) at diagnosis. Four patients without clinical follow-up were not included in these survival analyses. pos, positive; neg, negative.

Figure 4.

Molecular and clinical findings in patients with rare BRAF alterations. (A) Genome-wide coverage of fragments retrieved from a FFPE-TLC experiment targeting BRAF on a FFPE tissue sample from a patient from our cohort. A rearranged region to the BRAF gene (pink box) on chromosome 7 (Chr7) was identified by the concentration of fragments clustered around the BICD2 gene (yellow box) on chromosome 9. (B) Butterfly plot uncovering the reciprocal BICD2::BRAF rearrangement. Proximity-ligation products between the target gene (BRAF) and rearrangement partner (BICD2) are depicted (in green). Strand directions are indicated by arrows. See supplemental Figure 10 for details about FFPE-TLC technology. (C) Illustration of the identified BICD2::BRAF fusion. (D) Circos plot depicting the 3 distinct BRAF rearrangements identified in patients from our cohort (in orange), as well as other BRAF rearrangements previously identified in patients with LCH (in gray).^,,, , (E) Clinical and conventional radiography images of a patient from our cohort with BRAF^V600D-mutated SS-UFB LCH, who had a single osteolytic lesion in the right clavicle with remarkable abscess-like soft tissue extension through the skin. (F) Clinical and magnetic resonance imaging images of a patient from our cohort with BRAF^V600D-mutated SS-MFB LCH, who had a relapse of MFB disease with a remarkable orbital lesion with clear skin changes. (G) Coronal image of a chest computed tomography scan showing atypical pulmonary lesions in the patient with a BICD2::BRAF fusion. Shown are multiple solid nodules in both lungs, including a very large tumor in the right upper lobe measuring 55 × 18 × 15 mm. A biopsy of this tumor excluded cooccurrence of lymphoma or another disease and revealed clusters of CD1a⁺ CD207⁺ cells, compatible with LCH. (H) Coronal image of a chest computed tomography scan showing many large cystic lung lesions in a child with high-risk LCH harboring both BRAF^V600E and BRAF^R603Q mutations. Professional illustration of panel C made by ProteinPaint software. mb, megabase.

Figure 5.

Identified BRAF and MEK1 alterations. Schematic representations of BRAF (A) and MEK1 (B) proteins with alterations detected in patients from our cohort. MEK1 is encoded by the gene MAP2K1. Figures not entirely to scale. AS, activation segment; ATP, adenosine triphosphate binding site; C, C-terminus; CL, catalytic loop; CR, conserved region; D, docking; del, deletion; DFG, DFG motif; ins, insertion; N, N-terminus; NES, nuclear export signal; NRR, negative regulatory region; P, phosphate-binding; Pro, proline.