MR imaging biomarkers associated with patient-reported xerostomia post-RT

Proton therapy and NTCP models
Head-and-neck (H&N) cancer patients undergoing external beam radiotherapy with
photons often experience side-effects, such as xerostomia and dysphagia, that
significantly impair their quality of life. That happens because part of
healthy tissues, organs-at-risk (OARs) surrounding the tumor, such as the
salivary glands, are included in the irradiation field. The first observations
of radiation-induced toxicity are anatomical and morphological changes, such as
volume reduction, fibrosis, inflammation [4]. Also changes in cellularity are
observed. These objectively observed changes, that can be depicted and measured
by imaging, result in functional impairment and consequently to clinical
symptoms that can persist long after the end of the treatment. For example, the
functional impairment of the irradiated salivary glands is hyposalivation.
Depending on the severity of the hyposalivation patients experience clinical
symptoms such a dry mouth syndrome, dental carries, tooth decay, difficulties
chewing, speaking and swallowing.
Since 2019 it is possible in the Netherlands, to irradiate head-and-neck cancer
patients with proton beam therapy (PBT). PBT is expected to reduce toxicity
because less amount of healthy tissues will be involved in the irradiation
field [8]. However, the efficacy of this treatment, which is about 3 to 5 times
more expensive than the PBT, has not been proven yet [9-11]. Additionally,
because the capacity of the proton centers is limited, not all H&N cancer
patient can be treated by PBT. Therefore, the national head-and-neck
radiotherapy workgroup has defined selection criteria. The selection criteria
are based on normal tissue complication probability (NTCP) models that predict
xerostomia, dysphagia and tube feeding dependence at 6 months after the end of
radiotherapy (post-RT). In short, a photon to proton treatment planning
comparison is performed and only when patients have 10% less chance to one of
the above mentioned toxicities with PBT, they are considered eligible for PBT.
However, these NTCP models are not based on PBT data and have not been
validated for proton therapy or late toxicity yet . The national head-and-neck
radiotherapy workgroup, recognizes this and suggests that further NTCP model
validation should be the primary research agenda leading to urgently needed
evidence for proton therapy . Moreover, the current NTCP models are quite crude
and they do not take into account spatial information on the location of the
radiation damage within the salivary glands. Only the average dose to some of
the OARs is used as input. Due to advances in the treatment planning techniques
and for photon and for PBT therapy is possible to create steep dose gradients
within the OARs. Lately, it has been demonstrated that dose shape description
within the parotid glands can be beneficial for more accurate prediction of
xerostomia . Thus, quantifying and spatially localizing radiation-induced
damage to the healthy tissues surrounding the anatomically complex area of H&N
tumors, is important as input for future NTCP models for two main and equally
important reasons: (a) Improving treatment planning strategies for both photon
and proton therapy, with regard to steering dose gradients outside the tumor
volume, and prioritizing the order of OARs sparing. (b) Improving H&N cancer
patient selection for PBT contributes to the efficient use of the PBT capacity
in the Netherlands.

Radiation-induced damage to the salivary glands of head-and-neck (H&N) cancer
patients results in functional impairment and consequently in long-term
toxicity, xerostomia or dry mouth. Long term, late toxicity deteriorates
significantly the quality of life of these patients. Proton therapy has the
potential to reduce toxicity but it is 3-5 times more expensive than the photon
therapy and it is not available for all H&N cancer patients in the Netherlands.
Selection of patients for proton therapy is therefore needed. At the moment,
this is done based on toxicity prediction models developed for photon therapy.
However, these models are quite crude, are based on subjective measures of
functional impairment and do not take into account spatial information on the
location of the radiation-induced damage within the organs involved in the
salivary function. Our research aims to do that, by validating the MRI
biomarkers that are associated with patient-reported toxicity and that can
identify the location of radiation-induced damage.

This is an observational, cross-sectional prospective non-randomized two cohort
study. Two patient cohorts will be included in the study. Patients reporting
late toxicity of xerostomia (toxicity cohort) and patients with no xerostomia
complaints (non- toxicity cohort) post-RT.
Both patient cohorts undergo a one-time MRI exam during their follow-up period,
from 6>= months up to 3 years post-RT.

The potential risks for participating in this study are negligible. This study
can only be contacted to a group of H&N cancer patients because it is aiming to
validate MRI biomarkers that quantify radiation-induced xerostomia. The results
of this study will potentially benefit H&N cancer patients, treated by photon
RT or PBT in the future.
The patients treatment and treatment follow-up scheme are not altered because
of this study. However, the entire MRI scan will be reviewed by a radiologist,
who will inform the radiotherapist in case of incidental findings that may
need attention in terms of (altered) treatment.