Long-term outcome and prognosis of mixed histiocytosis (Erdheim-Chester disease and Langerhans Cell Histiocytosis).

Background: Erdheim-Chester disease (ECD) is a rare histiocytosis that may overlap with Langerhans Cell Histiocytosis (LCH). This "mixed" entity is poorly characterized. We here investigated the clinical phenotype, outcome, and prognostic factors of a large cohort of patients with mixed ECD-LCH.
Methods: This retrospective study was performed at two referral centers in France and Italy (Pitié-Salpêtrière Hospital, Paris; Meyer Children's Hospital, Florence). We included children and adults with ECD diagnosed in 2000-2022 who had biopsy-proven LCH, available data on clinical presentation, treatment and outcome, and a minimum follow-up of one year. Outcomes included differences in clinical presentation and survival between mixed ECD-LCH and isolated ECD; we also investigated response to treatments and predictors of survival in the mixed cohort. Survival was analyzed using the Kaplan-Maier method and differences in survival with the long-rank test. Cox regression models were used to evaluate the potential impact of age and gender on survival and to identify predictors of non-response and survival.
Findings: Out of a cohort of 502 ECD patients, 69 (14%) had mixed ECD-LCH. Compared to isolated ECD, mixed ECD-LCH occurred more frequently in females (51 vs. 26%, p  < 0.001) and in patients with multisystem disease (≥4 sites). Mixed ECD-LCH more frequently involved long bones (91 vs. 79%, p  = 0.014), central nervous system (51  vs. 34%, p  = 0.007), facial/orbit (52 vs. 38%, p  = 0.031), lungs (43 vs. 28%, p  = 0.009), hypothalamic/pituitary axis (51 vs. 26%, p  < 0.001), skin (61 vs. 29%, p  < 0.001), and lymph nodes (15 vs. 7%, p  = 0.028); the BRAF ^V600E mutation was also more frequent in mixed ECD-LCH (81 vs. 59%, p  < 0.001). Targeted treatments (BRAF and/or MEK inhibitors) induced response more frequently than conventional therapies (interferon-α, chemotherapy), either as first-line (77 vs. 29%, p  < 0.001) or as any line (75 vs. 24%, p  < 0.001). After a median follow-up of 71 months, 24 patients (35%) died. Survival probability was comparable between ECD alone and mixed ECD-LCH (log-rank p  = 0.948). At multivariable analysis, age at diagnosis (HR 1.052, 95% CI 1.008-1.096), associated hematologic conditions (HR 3.030, 95% CI 1.040-8.827), and treatment failure (HR 9.736, 95% CI 2.919-32.481) were associated with an increased risk of death, while lytic bone lesions with a lower risk (HR 0.116, 95% CI 0.031-0.432).
Interpretation: Mixed ECD-LCH is a multisystem disease driven by the BRAF ^V600E mutation and targeted treatments are effective. Age at diagnosis, bone lesion patterns, associated hematologic conditions, and treatment failure are the main predictors of death in mixed ECD-LCH.
Funding: None.
Competing Interests: AI reports research grants from Transgene, Sanofi, and Nutritheragene; consulting fees from Novocure, LeoPharma, Polytone Laser, and Novartis; honoraria from Novocure and Neurologies; travel funding from LeoPharma, Novocure, and Carthera. BT report consulting fees and honoraria from GSK, AstraZeneca, CSL Vifor, Boehringer Ingelheim, and Novartis; advisory board activity for Amgen.
(© 2024 The Author(s).)