Exploring Markov modeling approaches for the health economic assessment of circulating tumor cells in the management of metastatic prostate cancer patients

Objective: To investigate the applicability of methodological approaches in personalized (cancer) care, by using an early HTA case of the CTC trap medical device. Methods: For this study, different Markov cohort models were developed for the purpose of early HTA of the CTC-trap as diagnostic and/or prognostic device in the diagnostic pathway of prostate cancer. Results: The results demonstrate the difficulty of Markov modeling approaches in early HTA in personalized (cancer) care. Although, an indication that CTC-trap is a probable technology to use in the diagnostic pathway of prostate cancer, the evidence to substantiate this claim is difficult to obtain via Markov modeling. Due to multi-plural combination of treatments sequences, personalized medicine cannot apply the same simplifications and assumptions in straightforward intervention to golden standard comparisons. Conclusion: Two main problems were identified when Markov modeling approaches for early modeling of CTC-trap in mCRPC were modeled: time-dependency in transition probabilities and the available patient level data to populate the models in early HTA. The simplifications applied in health economic assessments cause major uncertainty and, especially in personalized medicine, may lead to large variations in outcomes. Keywords: Markov, Prostate Cancer, mCRPC, HTA, Decision modeling