Dose/Volume histogram patterns in Salivary Gland subvolumes influence xerostomia injury and recovery

Peijin Han, P. Lakshminarayanan, Wei Jiang, Ilya Shpitser, Xuan Hui, Sang Ho Lee, Zhi Cheng, Yue Guo, Russell H. Taylor, Sauleh Siddiqui, Michael R. Bowers, Khadija Sheikh, Ana P. Kiess, B.r. Page, Junghoon Lee, Harry Quon, Todd McNuttScientific 2019 · DOI

Abstract

Xerostomia is a common consequence of radiotherapy in head and neck cancer. This study compared regional radiation dose distributions in patients who developed xerostomia within 6 months of radiotherapy versus those who recovered within 18 months. Using a feature generation pipeline, dose-volume histogram (DVH) features (D10-D90) were extracted from geometrically defined subvolumes of the parotid glands (PG), submandibular glands (SMG), and oral cavity surrogates. Ridge logistic regression models were applied to predict injury and recovery. We found distinct regional dose patterns for each outcome: injury is primarily induced by subvolumes receiving higher doses, whereas recovery is preserved by reducing doses to subvolumes receiving lower doses. This suggests that sparing specific "low-dose" zones, particularly in the superior and middle PG, may preserve regenerative capacity.

Introduction

Radiotherapy for head and neck cancer (HNC) frequently causes xerostomia, a debilitating condition resulting from salivary gland damage. While current clinical guidelines use mean dose limits for the whole parotid gland, this approach does not account for regional functional differences within the glands. Some areas may be more prone to injury, while others may be essential for tissue regeneration. This research moves beyond whole-gland metrics to investigate how spatial dose distributions in partitioned sub-regions influence the trajectory of injury and subsequent recovery.

Methods

The researchers analyzed a cohort of 258 HNC patients treated between 2007 and 2017. The study defined "injury" as developing a xerostomia grade ≥ 2 within 6 months post-radiotherapy, and "recovery" as a subsequent reduction to grade < 2 by 18 months.

Patient selection flowchart and cohort breakdown for injury and recovery analysis.

Radio-morphologic Feature Pipeline

A technical pipeline was developed to partition the salivary glands into functional subvolumes:

  • Subvolume Partitioning: The parotid glands were divided into nine sub-regions based on radial sectors (medial, anterior, posterior) and superior-inferior sections.
  • Feature Extraction: 198 dosimetric features were extracted, focusing on D10 through D90 (the dose covering 10% to 90% of a subvolume).
  • Predictive Modeling: Ridge Logistic Regression was employed to handle high feature correlation, identifying the most significant predictors for injury and recovery.
Geometric partitioning of parotid (PG) and submandibular (SMG) glands into subvolumes.

Key Insights and Results

The analysis revealed that injury and recovery are decoupled processes driven by different dosimetric thresholds.

  • Injury vs. Recovery Thresholds: Injury (AUC 0.78) was primarily driven by high-dose "hot spots" (D10-D40). In contrast, recovery (AUC 0.70) was more sensitive to the "low-dose bath" (D70-D90). This suggests that even if a gland is injured, its ability to recover depends on sparing specific regions from lower radiation doses.
  • Regenerative Zones: The superior and middle portions of the parotid glands were identified as critical for recovery. These regions likely house essential structures for repair, such as main excretory ducts, stem/progenitor cells, and parasympathetic nerves.
  • Symmetry: Injury-related subvolumes tended to be symmetric across both sides of the face, while recovery was more dependent on specific unilateral sparing.

Conclusion and Future Directions

This study demonstrates that the spatial distribution of radiation dose is a critical factor in long-term salivary function. By sparing specific sub-regions associated with regeneration, clinicians may improve patient outcomes.

However, limitations remain. The study relied on physician-assessed CTCAE scales; future research should incorporate objective salivary flow measurements. Additionally, the biological mechanisms—specifically the exact location of human salivary stem cells—require further radiobiological validation. External validation across multiple centers will be necessary to confirm these findings for broader clinical application.

Summary updated

Abstract

Xerostomia is a common consequence of radiotherapy in head and neck cancer. The objective was to compare the regional radiation dose distribution in patients that developed xerostomia within 6 months of radiotherapy and those recovered from xerostomia within 18 months post-radiotherapy. We developed a feature generation pipeline to extract dose volume histogram features from geometrically defined ipsilateral/contralateral parotid glands, submandibular glands, and oral cavity surrogates for each patient. Permutation tests with multiple comparisons were performed to assess the dose difference between injury vs. non-injury and recovery vs. non-recovery. Ridge logistic regression models were applied to predict injury and recovery using clinical features along with dose features (D10-D90) of the subvolumes extracted from oral cavity and salivary gland contours + 3 mm peripheral shell. Model performances were assessed by the area under the receiver operating characteristic curve (AUC) using nested cross-validation. We found that different regional dose/volume metrics patterns exist for injury vs. recovery. Compared to injury, recovery has increased importance to the subvolumes receiving lower dose. Within the subvolumes, injury tends to have increased importance towards D10 from D90. This suggests that different threshold for xerostomia injury and recovery. Injury is induced by the subvolumes receiving higher dose, and the ability to recover can be preserved by further reducing the dose to subvolumes receiving lower dose.
BibTeX
@inproceedings{han2019dose,
  author = {Peijin Han and P. Lakshminarayanan and Wei Jiang and Ilya Shpitser and Xuan Hui and Sang Ho Lee and Zhi Cheng and Yue Guo and Russell H. Taylor and Sauleh Siddiqui and Michael R. Bowers and Khadija Sheikh and Ana P. Kiess and B.R. Page and Junghoon Lee and Harry Quon and Todd McNutt},
  title = {{Dose/Volume histogram patterns in Salivary Gland subvolumes influence xerostomia injury and recovery}},
  booktitle = {Scientific},
  year = {2019},
  doi = {10.1038/s41598-019-40228-y},
}
Figure