The efficacy of different follow-up strategies in clinical stage I Non-seminomatous Germ Cell Cancer: a Markov simulation study

Abstract

Objective: There is no universally accepted standard protocol for surveillance of patients with clinical stage I Non Seminomatous Germ Cell Tumors (CS I NSGCT). Prospective studies to compare different follow-up policies have not been performed, even though a great deal of time and resources is spent in surveillance. In this study, we constructed a Markov model to evaluate the impact of different follow-up strategies on disease-specific mortality (DSM) and life expectancy (LE) of patients with CS I NSGCT.

Methods: A discrete time non-homogeneous semi-Markov model was used to simulate different follow-up strategies for a hypothetical population of CS I NSGCT patients. Estimates of the model parameters were based on the literature. Output parameters were DSM and LE. Three different strategies were compared: (1) the intensive The Netherlands Cancer Institute/Antoni van Leeuwenhoek hospital (NCI/AvL) protocol; (2) the European Association of Urology (EAU) protocol; and (3) a hypothetical minimal protocol (i.e. follow-up limited to the first two years). Furthermore, we evaluated the impact of abdominal CT scans and chest X-rays on DSM.

Results: Comparing with the EAU protocol (DSM: 3.05%; LE: 53.3 years), the intensive NCI/AvL protocol leads to a 1.2% lower DSM and a 6 months higher LE (DSM: 1.81%; LE: 53.9 years). The hypothetical follow-up scenario during the first two years shows a DSM of 6.83% and an LE of 51.4 years. Abdominal CT scans of the retroperitoneal lymph nodes appear to be important, while chest X-rays have little impact on DSM.

Conclusion: A follow-up policy limited to the first two years will result in an unacceptable high percentage of death from disease (6.83%). The relatively small benefit of an intensive follow-up protocol as proposed by the NCI/AvL, compared to that of the EAU, must be weighed against its economic and psychological costs. Our model suggests that CT-scanning is essential for a low DSM, whereas the large number of X-rays seem to have little additional effect.