Purpose: Systemic mast cell disease (SMCD) follows an indolent course in most patients, but a significant number of patients die of neoplastic hematologic disorders. Reviews of the literature and retrospective studies in a single institution have defined features that may be associated with a poor prognosis, but prospective studies have been lacking. Therefore, we prospectively analyzed the relationship between clinical, laboratory, and hematopathologic findings and clinical outcome in a series of 46 patients with mast cell disease. This analysis was employed to both define clinically useful prognostic variables and describe the histologic evolution of bone marrow mast cell infiltration and its relationship to hematologic neoplasia.
Patients and methods: Forty-six adult patients were referred to the National Institutes of Health (NIH) with clinical and/or pathologic evidence of mast cell proliferation. All patients had bone marrow examinations, and 10 patients underwent serial bone marrow biopsies. The diagnosis of SMCD required pathologic documentation of bone marrow mast cell infiltrates. The patients were followed for up to 13 years at the NIH (up to 30 years after the initial pathologic diagnosis of mast cell disease). Statistical analysis defined the correlation between variables and the presence of diagnostic bone marrow lesions. The Kaplan-Meier method was used to construct survival curves, and the effects of various variables on the survival time were examined.
Results: Thirty-two of 46 patients (74%) had a bone marrow biopsy diagnostic for SMCD. The remaining 14 patients were considered to have cutaneous mast cell disease (CMCD). Univariate analysis showed that hepatosplenomegaly, alkaline phosphatase level, absolute lymphocyte count, and age at onset of symptoms were positively correlated with SMCD, whereas hemoglobin level was negatively associated with diagnostic bone marrow lesions. With multivariate analysis, only hemoglobin and absolute lymphocyte count remained as significant independent predictors of bone marrow findings. No CMCD patient died or had significant clinical deterioration in the 1- to 30-year period of follow-up (median = 8.5 years), whereas 10 of 32 SMCD patients (31%) died from 1 to 22 years after diagnosis (median = 2.5 years) (p less than 0.0001). Univariate analysis revealed the following variables as significantly increasing the risk of death in patients with SMCD: later onset of symptoms, absence of cutaneous mastocytosis, thrombocytopenia, elevated lactate dehydrogenase (LDH) level, anemia, bone marrow hypercellularity, qualitative peripheral blood smear abnormalities, elevated alkaline phosphatase level, and hepatosplenomegaly. Multivariate analysis showed that only the age at onset of symptoms and LDH levels were significant independent predictors of survival. Eight of the 10 SMCD patients who died had myeloproliferative or myelodysplastic syndromes or acute nonlymphocytic leukemia.
Conclusion: Our prospective study has defined a number of important variables in patients with clinical evidence of mast cell proliferation that can predict both the presence of SMCD and the likelihood of fatal disease. Since recent evidence suggests that mast cells derive from a bone marrow hematopoietic progenitor, SMCD may represent a myeloproliferative condition with the propensity to evolve into a neoplastic granulocytic disorder in a significant minority of patients.