BRAF^V600E is associated with higher incidence of second cancers in adults with Langerhans cell histiocytosis

TO THE EDITOR:

Langerhans cell histiocytosis (LCH) is a rare hematologic disorder with variable clinical presentations.¹ The discovery of somatic mutations in the mitogen-activated protein kinase (MAPK) signaling pathway established LCH as a clonal neoplasm.² In pediatric LCH, BRAF^V600E is detected in 50% to 60% of patients and associated with severe disease forms.^3,4 In adults, the prevalence and clinical associations of BRAF^V600E remain incompletely characterized. We present data from a large BRAF^V600E genotyped cohort of adults with LCH. In addition, we systematically reviewed published literature pertaining to frequency and clinical associations of BRAF^V600E in this population and provide a meta-analysis of relevant data.

We performed an observational cohort study of adults with LCH diagnosed and/or treated at Mayo Clinic or 1 of 3 Dutch academic medical centers. Ethical approval, or waiver of consent, was obtained from all Institutional Review Boards. Mutational analysis of LCH tissue samples was performed as part of routine diagnostics or in the context of this study using published methods.⁴ Techniques included BRAF^V600E allele-specific polymerase chain reaction (PCR), next-generation-sequencing (NGS), whole-exome sequencing, and/or BRAF-VE1 immunohistochemistry.^4,5 The latter was internally validated for the evaluation of histiocytoses by cross-testing with NGS.⁶ Disease extent was categorized into single-system (SS) and multisystem (MS) disease and subcategorized per consensus recommendations.¹ Events were defined as LCH progression, relapse, or death from any cause. Associations of BRAF^V600E with clinical characteristics were evaluated with Fisher exact or Mann-Whitney tests. Event-free survival (EFS) and overall survival were estimated by the Kaplan-Meier method and compared with log-rank tests. Cumulative incidences were calculated and compared by competing risk analysis, with death as a competing event using the Fine-Gray method. Standardized incidence ratios (SIRs) were calculated as the ratio of observed-to-expected numbers of second cancers, as previously described.⁷ Age-, sex-, and calendar year–specific cancer rates from The Netherlands Cancer Registry from 1989 to 2020 were used to calculate expected numbers; rates of 1989 were used for the period before 1989, and rates of 2020 were used for the period after 2020. Methods for systematic review and meta-analysis are detailed in the supplemental Methods (available on the Blood website).

We included 156 adults with LCH and known lesional BRAF^V600E status, diagnosed between 1984 and 2021 (85 from Mayo Clinic and 71 from The Netherlands). Patients from Mayo Clinic had more severe disease presentations and received chemotherapy more frequently (supplemental Table 1), likely representing referral bias. In the combined cohort, median age at diagnosis was 41 years (range, 18-88 years). Disease extent included 58 patients (37%) with MS-LCH, 22 patients (14%) with SS-pulmonary LCH, and 76 patients (49%) with nonpulmonary SS-LCH. BRAF^V600E was detected in 74 of 156 (47%) patients (Table 1 and supplemental Figure 1). In the BRAF^V600E-negative group, MAP2K1 mutations were detected in 20 of 41 cases (49%), BRAF exon 12 insertions/deletions were detected in 2 of 24 cases (8%), and a BRAF^R662K or a LMTK2:BRAF fusion was detected in single cases⁸ (supplemental Figure 1 and supplemental Table 2). BRAF^V600E did not correlate with age, sex, disease extent, sites of disease, or first-line therapy subtype (Table 1 and supplemental Figure 1). BRAF^V600E also did not correlate with reduced EFS (P = .63; Figure 1A) or overall survival (P = .23; Figure 1B), even when excluding patients who received targeted therapy (supplemental Figure 2). However, BRAF^V600E was associated with an increased incidence of second primary malignancies (SPMs) after LCH diagnosis, with a 5-year cumulative incidence of 17.3% (95% confidence interval [CI], 6.7%-27.9%) for BRAF^V600E-positive vs 4.1% (95% CI, 0.0%-8.7%) for BRAF^V600E-negative patients (Gray test P = .017; Figure 1C). This association was confirmed in multivariable analysis (supplemental Table 3). In addition, BRAF^V600E-positive but not BRAF^V600E-negative patients had a significantly elevated risk of SPMs when compared with the age-, sex-, and calendar year–matched general population of The Netherlands (SIR BRAF^V600E: 5.72; 95% CI, 3.13-9.60; SIR non-BRAF^V600E: 1.71; 95% CI, 0.47-4.37; supplemental Table 4). Of 18 patients with SPMs, 9 presented with second hematologic malignancies (SHMs), including 4 with acute myeloid leukemia (AML). Additional solid malignancies comprised many subtypes (supplemental Table 5). Both BRAF^V600E-positive and BRAF^V600E-negative patients had a significantly elevated risk of SHMs when compared with the age-, sex-, and calendar year–matched general population of The Netherlands (SIR BRAF^V600E: 32.71; 95% CI, 13.15-67.39; SIR non-BRAF^V600E: 9.54; 95% CI, 1.16-34.48; supplemental Table 4). Comparison of cumulative incidence of SHMs between BRAF^V600E-positive and BRAF^V600E-negative patients was not statistically significant (Gray test P = .08; Figure 1D), possibly due to lack of power. Notably, SHMs were not confined to patients with MS-LCH. Moreover, none of the patients with SHMs had prior exposure to alkylating agents or topoisomerase inhibitors, but all 4 with secondary AML had received cladribine (supplemental Table 6). When including malignancies occurring before LCH diagnosis, a higher frequency of additional malignancies in BRAF^V600E patients remained apparent (Figure 1E); particularly, the 2 patients with hematologic malignancies before LCH harbored BRAF^V600E (Figure 1F).

In case 2 with AML diagnosed 1 year after BRAF^V600E-negative MS-LCH, SRSF2^P95R was identified in both the LCH sample (variant allele frequency, 26%) and AML specimen (variant allele frequency, 49%). In addition, ASXL1, IDH2, and JAK2 alterations were detected in the AML specimen, but not in the LCH sample (Figure 1G). In case 8 with diffuse large B-cell lymphoma (DLBCL) diagnosed 3.5 years after SS-bone LCH (treated surgically), PCR–based IdentiClone analysis suggested identical immunoglobulin heavy chain (IGH) and immunoglobulin kappa light chain gene rearrangements in separate LCH and DLBCL samples (supplemental Figures 3 and 4). Additional NGS-based clonality assessment confirmed the presence of an identical IGH rearrangement (V3-48-1/14/-5 J4) in the DLBCL sample and microdissected LCH tissue devoid of B cells (supplemental Figure 5). In addition, distinct oncogenic genetic alterations unique to the LCH (BRAF^V600E) or DLBCL (BCL2 translocation; CARD11^S250P; CREBBP^R1446L; MYD88^S219C; and TP53^R282W) were detected (Figure 1H).

Regarding the meta-analysis (supplemental Figures 6 and 7), BRAF^V600E frequency in 712 evaluated adults with LCH was 38.9% (95% CI, 35.3%-42.5%). There was no correlation of BRAF^V600E with MS disease (P = .07; odds ratio, 0.75; 95% CI, 0.53-1.05) or with events (P = .14; odds ratio, 0.80; 95% CI, 0.54-1.18). No meta-analysis could be performed for SPMs, due to lack of reported data.

Our study represents the largest study to date focusing on frequency and clinical associations of BRAF^V600E in adults with LCH. BRAF^V600E frequencies in our cohort (47%) and meta-analysis (39%) were lower when compared with large studies of children with LCH.^3,4 Contrary to findings in pediatric LCH,^3,4 BRAF^V600E was not associated with clinical features or with inferior EFS. These findings are supported by prior reports^9-11 included in our meta-analysis. A limitation of our and other studies is that organ involvement and response to treatment were not uniformly assessed. Nevertheless, these data underscore that caution is warranted when adopting insights from pediatric LCH to the adult population.

Interestingly, BRAF^V600E correlated with a higher incidence of SPMs (Figure 1C and supplemental Table 4). A previous cohort study reported a high prevalence (32%) of additional malignancies in 132 adult patients with LCH.¹² Moreover, an increased risk of SPMs among patients with LCH was found in a recent population-based study,¹³ which also revealed that SPMs were the most common cause of death in adults with LCH. However, no genomic data for these patients were reported. Strikingly, 9 of 18 patients (50%) with SPMs in our cohort had SHMs (Figure 1F). This is in accordance with a recent study of LCH-associated malignancies.¹⁴ Furthermore, a high prevalence of myeloid malignancies has been reported in patients with Erdheim-Chester disease,¹⁵ a closely related histiocytic neoplasm also frequently characterized by BRAF^V600E. Although both BRAF^V600E-positive and BRAF^V600E-negative patients from our cohort had a significantly elevated risk of SHMs compared with the general Dutch population, patients with BRAF^V600E-positive LCH seem to be at highest risk of developing SHMs (Figure 1D). Given that LCH may arise from mutated bone marrow–derived progenitors with oligo- or multipotent differentiation capacity,⁵ SHMs may have shared clonal origins with LCH in some patients. Several cases with LCH and diverse hematologic malignancies harboring the same mutations, supporting this notion, have been reported.¹⁶ Our finding of identical genomic alterations in separate LCH and SHM samples (cases 2 and 8) suggests a shared clonal origin of both hematologic diseases and demonstrates the importance of analyzing such cases beyond BRAF^V600E. Notably, clonal immunoglobulin and T-cell receptor gene rearrangements have previously been detected in LCH cases, including patients without (known) SHMs,¹⁷ supporting the concept that histiocytoses may occasionally be derived from lymphoid-committed cells. Analysis of clonal relations in other cases from our cohort was not feasible due to unavailability of specimens. Prospective collection and genomic analysis of tissue, blood, and/or bone marrow samples of adult patients with LCH is warranted, as this will shed more light on associations between LCH and additional neoplasms.¹⁴ Finally, further studies are needed to identify prognostic factors beyond lesional BRAF^V600E status.

The authors thank the molecular diagnostics staff of the Department of Pathology of Leiden University Medical Center, as well as Michelle van der Klift, Fleur de Groot, Patty Jansen, and Karoly Szuhai for their help with genetic analyses. Sofie Breuking is acknowledged for prior work on the systematic review. The authors also thank Wichor Bramer from the Erasmus MC Medical Library for developing and updating the search strategies for the systematic review. Figure 1G-H was made with BioRender.com software.

This study was supported by grants from Stichting 1000 Kaarsjes voor Juultje to A.G.S.v.H., the University of Iowa/Mayo Clinic Lymphoma SPORE P50 CA97274 to G.G. and J.P.A., and the Walter B. Frommeyer, Jr, Fellowship Award in Investigative Medicine, University of Alabama at Birmingham, to G.G. P.G.K. received an MD/PhD grant from Leiden University Medical Center.

Contribution: C.v.d.B., J.A.M.v.L., R.S.G., G.G., and A.G.S.v.H. conceived the study; A.A.A.-M., P.G.K., T.C.E.Z., and J.P.A. collected clinical data; P.G.K., J.F.-B., E.C.S., S.D., R.A.L.d.G., A.W.L., T.v.W., and A.G.S.v.H. performed additional molecular genetic analyses outside routine diagnostics; P.G.K. analyzed the combined clinicogenomic data set and drafted the (supplemental) figures and tables; T.C.E.Z. performed the systematic review and meta-analysis, with P.G.K. as second assessor; E.A.M.F. supervised the systematic review and statistical analyses; J.C.T. helped with calculating standardized incidence ratios for second cancers; A.A.A.-M., P.G.K., T.C.E.Z., R.S.G., G.G., and A.G.S.v.H. drafted the manuscript; A.A.A.-M., J.P.A., N.N.B., S.M.S., M.V.S., C.D.-P., M.J.K., J.H.R., R.V., W.O.T., J.R.Y., K.R., A.R., A.H.G.C., R.M.V., C.J.M.v.N., B.V.B., G.J.B., P.S., J.A.M.B., J.S.P.V., M.A.J.v.d.S., E.F.S., S.H.T., C.v.d.B., J.A.M.v.L., R.S.G., and G.G. were involved in clinical care for included patients; and all authors critically reviewed and approved the manuscript.

Conflict-of-interest disclosure: G.G. served on an advisory board for Opna Bio LLC, received royalties from UpToDate, and is a consultant to 2nd.MD. W.O.T. has received research funding from Mallinckrodt Inc, the Center for Multiple Sclerosis and Autoimmune Neurology at Mayo Clinic, and the National Institutes of Health; speaking fees from DKBMed, NeurologyLive, and the American Association for Clinical Chemistry; and publishing royalties from the publication of Mayo Clinic Cases in Neuroimmunology. The remaining authors declare no competing financial interests.

Correspondence: Astrid G.S. van Halteren, Department of Internal Medicine/Clinical Immunology, NFU Center for Histiocytic Disorders, Erasmus University Medical Center, PO Box 2040, 3000 CA, Rotterdam, The Netherlands; e-mail: a.vanhalteren@erasmusmc.nl; and Gaurav Goyal, Division of Hematology-Oncology, University of Alabama at Birmingham, 1802 6th Ave S, Suite 2555 NP, Birmingham, AL 35294; e-mail: ggoyal@uabmc.edu.