# Neurovascular bundle sparing in hypofractionated radiotherapy maintained with realistic treatment uncertainties

## TL;DR

This study shows that sparing neurovascular bundles in prostate radiotherapy is feasible even with treatment uncertainties, without compromising target coverage.

## Contribution

A novel use of polynomial chaos expansion to simulate treatment uncertainties in neurovascular bundle sparing during hypofractionated radiotherapy.

## Key findings

- Sparing neurovascular bundles reduced median dose by 3.7 Gy and 4.4 Gy in non-coplanar and coplanar plans.
- Clinical target volume dose remained adequate despite neurovascular bundle sparing.
- Treatment uncertainties had limited probability of exceeding planning constraints for organs at risk.

## Abstract

•Neurovascular bundle sparing maintained with realistic treatment uncertainties.•Adequate clinical target volume coverage with neurovascular bundle sparing.•Realistic simulation of treatment uncertainty using polynomial chaos expansion.•Statistically accurate impact of treatment uncertainties in hypofractionation.

Neurovascular bundle sparing maintained with realistic treatment uncertainties.

Adequate clinical target volume coverage with neurovascular bundle sparing.

Realistic simulation of treatment uncertainty using polynomial chaos expansion.

Statistically accurate impact of treatment uncertainties in hypofractionation.

Erectile dysfunction is a common side effect of radiotherapy for prostate cancer. To mitigate this toxicity, it has been suggested to limit the dose to critical nerves and vessels. We investigated the feasibility of sparing the neuro-vascular bundles (NVBs) in stereotactic body radiotherapy under the impact of realistic treatment uncertainties.

Non-sparing and sparing NVB treatment plans, delivered in 5 × 7.25 Gy, were automatically generated for 20 patients. Polynomial Chaos Expansion (PCE) was used to fast and accurately model the dose against treatment errors. PCE enabled a robustness evaluation of 100.000 treatment scenarios per plan, allowing to derive scenario distributions of clinically relevant dose volume histogram parameters and population dose histograms.

An average decrease of 3.7 Gy and 4.4 Gy in the median D0.1cm3 of the NVB was achieved in the patient population in the presence of realistic treatment uncertainties for non-coplanar (NC) and coplanar (C) plans respectively. Sparing NVBs decreased planning target volume coverage by 2.1 % in V36.25Gy on average, however clinical target volume (CTV) dose remained adequate. Population dose histograms showed that, while sparing does impact dose volume histogram parameters of organs at risk (OARs), the probability of a scenario exceeding planning constraints was limited.

NVB sparing was maintained in the presence of treatment uncertainties without compromising CTV coverage or OAR dose. There was no significant difference in the achieved NVB dose between NC and C plans. The clinical impact of the achieved sparing is subject of ongoing clinical trials.