From January 2010 to December 2014, 150 children with MS-LCH were treated in our hospital following a LCH II (Arm B) based protocol. Treatment was based on a modification of the LCH-II (Arm B) based protocol. However, the continuation treatment was extended to 56 weeks and etoposide was omitted from the continuation treatment. For the 59 patients with RO involvement (RO+) (the lungs are not considered a RO in the current study), the rapid response rate (week 6) was 61.0% and the 3-year overall survival (OS) 73.4±5.9%. Rapid responders had a better 3-year survival rate than poor responders (90.9±5.0% vs. 45.7±11.0%, P\<0.001). The 3-year OS in the current study is 10\ 20% lower than the rates reported by Gadner et al. and Morimoto et al.. We have not yet adopted effective salvage therapies for RO+ patients with recurrent disease. During the time of this study, cladribine was unavailable. Second-line therapy for non-responders or patients with disease reactivation was individualized treatment based on the physician's experience. An effective salvage therapy is essential for this high-risk group. For 91without RO involvement (RO-), 78 patients (85.7%) were rapid responders at week 6. The 3-year cumulative reactivation rate was 10.7% for RO- patients. No death occurred in this subgroup, with a 3-year OS of 100% in RO- patients. Compared to the LCH II and LCH III trials, the current study had a more intensive initial treatment regimen for RO- patients. However, the addition of etoposide to prednisone and vincristine in the initial therapy did not increase the 6-week response rate for RO- patients (85.7% in this study compared to 83% in the LCH II study and 86% in the LCH III study). Surprisingly, with a relatively intense initial treatment, a relatively low 3-year cumulative reactivation rate was observed in RO- patients in the current study. This result suggests that the initial treatment intensity and duration of continuation therapy both impact disease reactivation. The intensity of induction can affect the degree of disease resolution. Insufficient treatment intensity might lead to late relapse. Similarity to that observed has been in other childhood hematological malignancies. This finding deserves to be tested in prospective clinical trials with long-term follow-up. Cytarabine has been applied for patients with LCH but has never been evaluated in our hospital prospectively. In this study, we administer a cytarabine contained protocol to patients with multisystem involvement with or without risk organs involvement. The treatment results will be compared with our historical studies.