Poor Pulmonary Function is not Associated with Increased Rates of Toxicity or Decreased Overall Survival after Stereotactic Body Radiotherapy for Early Stage Non-small Cell Lung Cancer: Results of a Multi-Institutional Analysis

Reviewer: Lara Bonner Millar, MD
The Abramson Cancer Center of the University of Pennsylvania
Last Modified: October 31, 2010

Authors: M. Guckenberger1, J. Belderbos2, A. Hope3, L. L. Kestin4, M. Werner-Wasik5, J. J. Sonke2, Y. Xiao5, D. Yan4, J. Wilbert1,I. S. Grills4

1Julius-Maximilians University, Wuerzburg, Germany
2The Netherlands Cancer Institute, Amsterdam, Netherlands
3PrincessMargaret Hospital and University of Toronto, Toronto, ON, Canada
4William Beaumont Hospital, Royal Oak, MI
5Thomas Jefferson University, Philadelphia, PA


  • The standard treatment for early stage lung cancer is surgical resection
  • Some patients have medical co-morbidities that make them poor surgical candidates; in these "medically inoperable" patients, radiation treatment traditionally consisted of standard fractionation to a dose around 66 Gy. This treatment has been previously demonstrated to yield 5 year overall survival in the range of 15%.
  • Recently, stereotactic body radiotherapy (SBRT) has emerged as the treatment of choice for medically inoperable patients with early-stage lung cancer , based on phase I and phase II trials, which have shown local control of 85-95% and overall survival of 40-75% at two to three years
  • Based on earlier studies of SBRT, certain tumor factors, such as close proximity to the tracheobronchial tree and chest wall, have been found to be predictive of toxicity from treatment
  • Pulmonary function testing is important in the assessment of a patient’s candidacy for surgery for treatment of lung cancer, but the significance of pulmonary function results has not been well described in patients who are treated with SBRT. This study looked at pre-treatment pulmonary function to examine its role in predicting for survival and toxicity after SBRT.

Materials and Methods

  • This was an international, multi-institutional study of 411 patients (434 targets) with stage cT1-cT3 cN0 NSCLC treated with SBRT between 1998 and 2009 at five radiotherapy departments.
  • Median patient age was 74 years (range: 42-92)
  • Median tumor diameter was 2.4 cm with a maximum tumor size of 8.5 cm.
  • Various fractionation schemes were used, with a median of 3 treatment fractions delivered (range: 1 to 15)
  • Median total dose was 54 Gy (range: 20 Gy to 64 Gy).
  • All treatments used daily image-guidance
  • Only 63% had biopsy proven lung cancer; the biopsy rate was higher in Canada and the USA as compared to Europe
  • Most were PET staged
  • Median follow-up was 12 months; follow-up was greater than 2 years for 23% of the patients.


  • Pre-treatment pulmonary function tests (PFTs) data availability :
    • 83% had FEV1 absolute and predicted
    • 65% had DLCO predicted
    • 37% had actual DLCO actual
  • The median time between PFTs and start of SBRT was 35 days.
  • The median pre-treatment actual FEV1 was 1.4l (range 0.43l to 4.4l)
  • The Percent Predicted FEV1 was 65% (range 21% to 286%).
  • The median actual DLCO was 11.7 CO/mm Hg/min (range 3.4 CO/mm Hg/min to 24.2 CO/mm Hg/min)
  • The median predicted DLCO was 53% (range10% to 103%).
  • Pre-treatment pulmonary function was significantly different between the five institutions: median FEV1 was 1.6l and 1.2l at the institutions with the best and worst pulmonary function. Pre-treatment PFTs tended to be better in the USA and Canada and worse in Europe.
  • Two-year overall survival (OS) was 64%
  • The rate of pneumonitis grade II or higher was 6%. There was no correlation between pre-treatment PFTs and grade II or higher pneumonitis
  • Using the log-rank test, actual and predicted DLCO and actual FEV1 were significantly correlated with OS.
  • Two-year OS was 47% vs 74% for patients with DLCO maximum less than 9.7 CO/mm Hg/min vs greater than 9.7 CO/mm Hg/min, respectively (p = 0.002)
  • 2 year OS was 44% vs 64% for patients with FEV1 max less than 0.95l vs more than 0.95l (p = 0.01).
  • In the multivariate analysis, the relationship between PFTs and overall survival was not statistically significant. PFTs also were not predictive of CSS.

Author's Conclusions

  • PFTs differed significantly between the five institutions
  • Poor pulmonary function did not result in increased rates of pneumonitis after SBRT and should not be a contraindication for SBRT.
  • The influence of poor PFTs on overall survival is unclear.

Clinical Implications

  • Our understanding of the influence of tumor factors (size, location, etc) is better understood than the patient factors that contribute to prognosis and toxicity after SBRT.
  • This study suggests that not all medically inoperable patients are alike, however, the data for overall survival according to PFTs did not remain significant in multivariate analysis
  • The study did not assess smoking status and the impact on pneumonitis, but this may factor into toxicity as well as influence the baseline PFTs
  • Pre-treat PFT results were better in USA and Canada as compared to Europe, suggesting different geographic thresholds for determining patients to be "medically inoperable"  
  • Overall, this study did not demonstrate poor pulmonary function to influence outcomes after SBRT on multivariate analysis, and poor PFTs should not be considered a contraindication to SBRT for medially inoperable patients. Further work may allow more clear understanding of the contribution of overall lung function to outcomes after SBRT.

Frequently Asked Questions