Most cost-effectiveness models for evaluation of cancer care compare interventions within a single treatment line. However, to investigate the full impact of a new treatment, also downstream effects must be taken into account. Furthermore, most models are based on observed clinical states, whilst these observations depend on the timing of examinations and the choice of diagnostic test. To evaluate the potential of new treatments and diagnostics, the underlying disease process needs to be modeled including the interaction with diagnostics and treatment. OBJECTIVES: To build a flexible framework for a disease model, that simulates cancer progression to obtain clinical, patient and economic outcomes, while taking diagnostics treatment pathways and surveillance schedules into account . METHODS: The modeling framework discerns two levels to describe disease progression, the level of the patient and the tumor. At the patient level, an individual is characterized by clinical states; “primary tumor only”, “local recurrence”, “regional recurrence”, “distant metastasis, stable”, “distant metastasis, progressing” and “death”. The clinical state is derived from disease development at the tumor level.  Seven tumor growth states are defined: “absent tumor”, “dormant tumor”, “micro tumor”, “small macro tumor”, “medium macro tumor”, “large macro tumor”, “symptomatic tumor”. Melanoma progression was used as a case study. The model simulates, in parallel, 11 possible tumor sites, ranging from “local” to “regional” and “distant metastatic” locations. Sites were chosen because they are associated with different treatment and prognosis.   The disease model is complemented with a treatment and surveillance module. In this module, treatment choices in each of the clinical states are specified. Treatment choice may depend on patient and tumor features, and subsequently influences rate of transitioning between tumor growth states. For surveillance, timing of surveillance visits, techniques used and their detection rate(s) are specified. CONCLUSIONS: The proposed framework provides a flexible and widely applicable cancer modeling design.