# Evaluation of acute ocular toxicity after definitive-intent radiation therapy in canine sinonasal tumors

## TL;DR

This study evaluates acute eye damage in dogs treated for nasal tumors with radiation, finding that most cases had mild effects and new guidelines can help reduce risks.

## Contribution

The study introduces new dose-response thresholds and validates practical toxicity scoring systems for canine ocular radiation therapy.

## Key findings

- 20% of eyes showed clinically relevant ocular toxicity using the modified McDonald-Shadduck system.
- Adherence to institutional dose constraints led to less than 2% clinically relevant toxicity using VRTOG scoring systems.
- VRTOG2.0 accurately reflects detailed toxicity assessments while being more practical for routine use.

## Abstract

Acute toxicity and survival were documented in dogs undergoing radiotherapy for sinonasal tumors. However, few studies specifically investigated ocular toxicity, resulting in limited data on radiation tolerance thresholds. This study aimed to quantify acute ocular toxicity, compare toxicity levels using three different scoring systems, and establish practical preliminary dose-response limits. Dogs with sinonasal tumors treated with a definitive-intent 10-fraction radiotherapy protocol were included if they underwent prospective ophthalmic examinations, both prior to radiotherapy and at least once within three months following treatment. Ocular toxicity was assessed using three scoring systems: 1) Modified McDonald-Shadduck (evaluated by ophthalmologists), 2) Veterinary Radiation Therapy Oncology Group (VRTOG) 1.0 (evaluated by radiation oncologists), and 3) VRTOG2.0 (evaluated retrospectively). Radiation doses to each eye, retina, cornea, lacrimal, and accessory lacrimal gland were documented, and adherence to the new institutional ocular dose constraints was analyzed. Seventy client-owned dogs (140 eyes) were enrolled between 2016 and 2024, yielding 241 ophthalmic examinations. Clinically relevant ocular toxicity was identified in 28 eyes (20%) according to the modified McDonald-Shadduck system, 14 eyes (10%) using the VRTOG1.0 system (grade ≥2), and in 15 eyes (11%) using VRTOG2.0 toxicity (grade ≥3). Adherence to our institutional constraints resulted in low rates of clinically relevant toxicity: only 0.7% of eyes according to VRTOG 1.0 and 1.4% according to VRTOG2.0. TD5% and TD50% risk thresholds were established to facilitate optimized planning and reduced ocular risks. In conclusion, all dogs showed rather mild and only rarely clinically relevant ocular toxicity. Clinically relevant results of the very detailed modified McDonald-Shadduck scoring system seem to be accurately reflected by the VRTOG2.0 scoring system, which is more practical for daily clinical assessments. Detailed ocular evaluations are recommended primarily for dogs with clinically relevant VRTOG2.0 toxicity. Adherence to institutional ocular dose constraints significantly minimizes the risk of clinically relevant ocular toxicity.