KEY POINTS
- This retrospective single-center study from Princess Margaret Cancer Centre, Toronto, evaluated 280 patients receiving palliative thoracic radiotherapy between January 2015 and June 2023. Eligible patients had diagnostic chest CT within three months before RT, with palliative prescriptions limited to ≤30 Gy in ≤10 fractions. Median age was 66 years, and primary malignancies included lung cancer in 128 patients (45.7%), head and neck cancer in 34 (12.1%), and other cancers in 118 (42.1%). The most common treatment regimen was 20 Gy in five fractions (68.2%).
- A semiautomated three-dimensional CT reconstruction was used to assess central airway obstruction. Airway segmentation was performed using commercially available image-processing software, followed by blinded review by a thoracic radiologist. Nine central airways, including the trachea, main bronchi, bronchus intermedius, and lobar bronchi, were classified as patent or completely obstructed. Partial narrowing was not quantified because the three-dimensional models could not reliably assess luminal stenosis severity.
- Complete obstruction of at least one central airway was identified in 106/280 patients (37.9%), while 174 (62.1%) had all evaluated airways patent. Among patients with obstruction, 62 had one obstructed airway, 11 had two, 20 had three, and 13 had four or more. Clinical airway events were documented in 136 patients (48.6%), including 73 catastrophic events, defined as grade ≥3 obstruction or hemorrhage requiring major or urgent intervention, and 63 non-catastrophic events. These clinical events were assessed in the pretreatment setting, not through prospective surveillance after RT.
- Any CT-detected airway obstruction was associated with significantly higher odds of clinical airway events. After adjustment for age, sex, ECOG performance status, cancer site, and stage, patients with obstruction had 2.60-fold higher odds of an airway event (95% CI 1.56–4.40) compared with patients whose central airways remained patent. However, binary classification of any obstruction did not independently demonstrate higher odds of catastrophic events (OR 1.44, 95% CI 0.81–2.53), indicating that simply identifying one completely obstructed airway was insufficient to characterize the risk of severe clinical deterioration.
- The number and anatomical location of obstructed airways provided additional information about event severity. Each additional obstructed airway was associated with increased odds of any clinical airway event (adjusted OR 1.48, 95% CI 1.20–1.88) and catastrophic airway events (adjusted OR 1.31, 95% CI 1.07–1.61). Obstruction of the right or left main bronchus was particularly relevant, with adjusted OR 3.25 (95% CI 1.23–9.63) for any clinical airway event and OR 3.77 (95% CI 1.45–9.94) for catastrophic events.
- Airway obstruction was associated with shorter unadjusted survival, but not independently after adjustment. Median overall survival was 144 days (95% CI 120–200) among patients with obstructed airways versus 226 days (95% CI 183–304) among those with patent airways. The unadjusted mortality hazard ratio was 1.43 (95% CI 1.09–1.87), decreasing to 1.29 (95% CI 0.97–1.72) after multivariable adjustment. The number of obstructed airways was also not independently associated with mortality (adjusted HR 1.06, 95% CI 0.96–1.17). Overall cohort median survival was 194 days, with one-year survival of 29.5%.
- The study establishes association rather than prospective prediction of impending airway emergencies. Many symptomatic events, including indications for urgent radiotherapy, were already present when the pretreatment CT and consultation data were evaluated. Only 280 patients met inclusion criteria from a much larger palliative RT population, and airway reconstructions were reviewed by a single radiologist without interobserver reliability assessment. The analysis did not evaluate incomplete airway narrowing, dynamic obstruction, bronchoscopy findings, or whether earlier image-based intervention actually prevents hemorrhage, hospitalization, or death.
CLINICAL TAKEAWAY
Complete central airway obstruction visible on routine CT is associated with clinically important airway events among patients referred for palliative thoracic radiotherapy, with the number and location of obstructed bronchi providing additional information about event severity. Structured review of central airway anatomy may help prioritize multidisciplinary assessment and intervention, but prospective studies are needed to establish whether this approach predicts future events or improves clinical outcomes.