Chalmers Conferences, 9th European Conference on Mathematical and Theoretical Biology

Linear dependency between residual tumor fraction and regression probability can provide decision support after partial resection of pilocytic astrocytoma
Thomas Buder

Last modified: 2014-06-09

Abstract


Pilocytic Astrocytoma (PA) is a grade I brain tumor which is common in children. These
tumors are often benign and have a good prognosis, but in some cases a more aggressive
form is observed causing higher mortality. If possible, total resection is the treatment of
choice and provides a good prognosis. In many cases, only partial resection is possible due to
the location of the tumor. Then, the prognosis is largely unpredictable and there is
controversy about further treatment and the required follow‐up investigations. In some
cases of partial resection, the tumors spontaneously regress, whereas in other cases they
regrow or even progress to a more aggressive form. The dependency between residual
tumor size and regression probability is not understood yet. However, it could provide
clinicians decision support in two ways. First, it could help to decide whether a "wait and
see" strategy or, alternatively, additional therapy with further risks is required after partial
resection. Second, it could justify the risk of additional resections in certain cases if this
would imply a significant increase of the regression probability. A mathematical model is
introduced in order to investigate this dependency. The dynamics of the model is entirely
determined by a risk coefficient ɣ. This parameter is obtained by incorporating
epidemiological data about PA. It turns out that the dependency between residual tumor
fraction and regression probability is approximately linear. These insights can support
clinicians in their decision to adopt a wait and see strategy and to balance between the risk
of operation or further therapies and the risks of possible regrowth and progression of PA.