We introduce MASON, a fast, easily extensible, discrete-event multi-agent simulation toolkit in Java. MASON was designed to serve as the basis for a wide range of multiagent simulation tasks ranging from swarm robotics to machine learning to social complexity environments. MASON carefully delineates between model and visualization, allowing models to be dynamically detached from or attached to visualizers, and to change platforms mid-run. We describe the MASON system, its motivation, and its basic architectural design. We then compare MASON to related multi-agent libraries in the public domain, and discuss six applications of the system we have built over the past year to suggest its breadth of utility.

Fifty-two of 175 (30%) survivors of allogeneic marrow transplantation developed chronic graft-versus-hose diseases (GVHD). Five with limited chronic GVHD had an indolent clinical course with involvement of only the skin and liver. Forty-seven with extensive chronic GVHD had an unfavorable multiorgan disorder that resembled several autoimmune diseases. Thirteen patients with extensive disease (group I) were not treated and only 2 survive with Karnofsky scores >- 70%. Mortality resulted from infections and morbidity from sica syndrome, pulmonary and hepatic insufficiency, scleroderma-like skin disease, and contractures. Another 13 (group II) received a median of 8 mo prednisone and/or a brief course of antithymocyte globulin, and 3 survive without disability. The other 21 (group III) were treated with a combination of prednisone (1.0 mg/kg/q.o.d.) and either cyclophosphamide, procarbazine, or azathioprine (all 1.5 mg/kg/day) for a median of 13 mo. Combination therapy was well tolerated with only modest myelotoxicity. Fifteen in group III had a good and 4 a fair response to treatment while 2 with no response died. Azathioprine and prednisone was the most effective regimen. All therapy has been discontinued in 12 group III patients: GVHD returned in 5 (including 2 who died in spite of retreatment) while 7 remain free of GVHD for a median of 11 (range 6–30) mo observation. Only I group III survivor is disabled and 16 of the original 21 are alive 2–4 yr after transplant with Karnofsky scores of 70%-100%. Thus, combination immmunosuppression appears to favorably affect and, in some cases, premanently arrest the adverse natural course of extensive chronic GVHD.

The influence of busulfan (BU) plasma concentration on outcome of transplantation from HLA identical family members for the treatment of chronic myelogenous leukemia (CML) was examined in 45 patients transplanted in chronic phase (CP) (n = 39) or accelerated phase (AP) (n = 6). All patients received the same regimen of BU, 16 mg/kg orally and cyclophosphamide (CY), 120 mg/kg intravenously. Plasma concentrations of BU at steady state (CSSBU) during the dosing interval were measured for each patient. The mean CSSBU was 917 ng/mL (range, 642 to 1,749; median, 917; standard deviation, 213). Of patients with CSSBU below the median, seven (five of 18 in CP and two of four in AP) developed persistent cytogenetic relapse and three of these patients died. There were no relapses in patients with CSSBU above the median. The difference in the cumulative incidence of relapse between the two groups was statistically significant (P = .0003). CSSBU was the only statistically significant determinant of relapse in univariable or multivariable analysis. The 3-year survival estimates were 0.82 and 0.64 for patients with CSSBU above and below the median (P = .33). There was no statistically significant association of CSSBU with survival or nonrelapse mortality, although the power to detect a difference in survival between 0.82 and 0.64 was only 0.24, similarly CSSBU above the median was not associated with an increased risk of severe regimen-related toxicity. We conclude that low BU plasma levels are associated with an increased risk of relapse.