Radiation Modality is a Strong Predictor for Postoperative Pulmonary and GI Complications after Trimodality Therapy for Esophageal Cancer

Reporter: Annemarie Fernandes, MD
The Abramson Cancer Center of the University of Pennsylvania
Last Modified: October 29, 2012

Presenter: J. Wang, MD
Presenter's Affiliation: The University of Texas, MD Anderson Cancer Center


  • Esophageal cancer is the sixth leading cause of death from cancer worldwide. Operable patients are typically treated with neoadjuvant chemoradiation followed by surgical resection
  • The authors note that about 50% patients experience severe post-operative complications, the risk of which may be increased from radiation-induced toxicity pre-operatively.
  • Advanced radiation modalities such as Intensity Modulated Radiation Therapy (IMRT) or Proton Beam Therapy (PBT) have theoretical dosimetric advantages over conventional 3D-conformal radiation therapy (3D-CRT), but the clinical evidence is lacking
  • The purpose of this study was to evaluate post-treatment and post-operative complications in patients treated with these 3 different radiation treatment modalities

Materials and Methods

  • This is a retrospective review of 444 patients were treated with 3D-CRT (N=208, 1998-2008), IMRT (N=164, 2004-2011) or PBT (N=72, 2006-2011)
  • Patients were treated with neoadjuvant chemoradiation followed by surgical resection
  • Locally advanced, non-metastatic patients and patients surviving at least 30 days post-operatively were included in this study


  • Baseline patient and tumor characteristics were relatively balanced between the two groups, and 90% of patients had adenocarcinoma. About half of the patients received induction chemotherapy
  • Radiation modality was not significantly associated with post-operative cardiac (MI, AF and CHF) or wound complications
  • The rate of pulmonary complications (ARDS, pleural effusion, respiratory insufficiency, and pneumonia) was 25% and the rate of GI complications (leak, ileus, fistula) was 23%
  • There was a significant reduction in postoperative pulmonary complications for IMRT compared to 3D-CRT (OR: 0.5) and for PBT compared to 3D-CRT (OR: 0.32), but not for IMRT compared to PBT, after adjusting for pre-RT DLCO.
  • Post-operative GI complications were reduced in patients treated with for IMRT compared to 3D-CRT (OR: 0.57) and not significantly reduced for PBT compared to 3D-CRT (OR: 0.56). There was also no significant difference between IMRT and PBT.
  • The median hospital stay was significantly different between 3D-CRT (12 days) vs. IMRT (10 days) vs. PBT (8 days), p<0.001.
  • Dosimetric factors including mean lung dose (MLD) and mean heart dose (MHD) were significantly different between the 3 modalities. When MLD and MHD were included in the multivariable analysis evaluating pulmonary complications, radiation modality was no longer associated with pulmonary morbidity. Only MLD correlated with pulmonary complications
  • The authors presented a scatter plot with pulmonary complications on the y-axis and mean lung dose on the x-axis. On this plot, the proton beam patients appeared to have the lowest MLD parameters, which correlated with lower risk of pulmonary toxicity.

Author's Conclusions

  • The authors conclude that IMRT and PBT significantly reduced postoperative pulmonary and GI complication rates compared to 3D-CRT in esophageal cancer patients.
  • The authors attribute the difference in pulmonary complications to dosimetric differences between the 3 modalities, particularly MLD.

Clinical Implications

  • Esophageal cancer is the sixth leading cause of death from cancer worldwide. Neoadjuvant chemoradiation followed by surgical resection has been shown to improve overall survival compared to surgical resection alone. Neoadjuvant therapy with chemotherapy and radiation therapy, however, can increase treatment-related toxicity. Decreasing radiation toxicity by optimizing radiation therapy targeting can potentially improve adverse effects.
  • It is not surprising that IMRT had advantages over 3D-CRT. 3D-CRT does not modulate or optimize the intensity of the beam and is delivered using similar beam arrangements as IMRT.
  • Proton beam therapy is a newer radiation technique that allows sparing of normal tissue and has the potential to reduce toxicity. Proton beam therapy is expensive and randomized data demonstrating improved clinical outcomes for esophageal patients have not been previously reported.
  • Evaluating different treatment modalities is dependent on patient anatomy. Esophageal cancers occur at different locations throughout the esophagus. Treatment of mid and lower esophageal cancers likely results in increased cardiac morbidity as the heart is more likely to be in the radiation treatment field. In these cases, proton therapy may have an advantage as posteriorly oriented proton therapy beams can potentially spare radiation dose to the heart. This study did not find any differences in cardiac morbidity. However, they evaluated 72 patients treated with proton therapy and did not stratify by location of tumor. Cardiac morbidity if often a late complication of treatment and longer follow-up is needed.
  • The authors found that PBT improves pulmonary complications when compared to 3DCRT. There was no statistically significant advantage over IMRT, however. PBT did not improve the rate of GI complications complications (leak, ileus, fistula) compared to IMRT or 3D-CRT.
  • There was a trend towards decreased lung dosimetric parameters with proton therapy, however the study was likely underpowered to detect a statistical difference in dosimetric data or clinical outcomes.
  • Proton beam therapy is generally delivered with fewer beams compared to IMRT. The authors generally used 2 beams for proton therapy and at least 4-5 beams for IMRT. While proton beam therapy did not show an advantage to IMRT in this acute to subacute setting, the potential advantage of proton beam therapy may be related to lower integral dose to the patient, resulting in lower long-term toxicity.
  • Additionally, other lung parameters, such as the V5 (volume of lung receiving 5 Gy), have been shown to correlate with the development of radiation pneumonitis. It is likely that proton beam therapy, by using fewer beams, has lower V5 values.
  • Patients treated with IMRT and PBT had shorter hospital stays, but this may be reflective of the treatment time as patients treated with IMRT and PBT were treated more recently and post-operative, supportive care has improved over time.
  • Evaluating post-operative mortality in esophageal patients treated with neoadjuvant chemoradiation is important as the morbidity of neoadjuvant therapy may contribute to post-operative mortality. This study excluded patients who did not survive more than 30 days after surgery. It would be interesting to study correlations between radiation treatment modality and post-operative mortality.
  • Prospective randomized control trials are necessary to evaluate the impact of different radiation treatment modalities on clinical outcomes in patients treated with neoadjuvant chemoradiation for esophageal cancer; however, this study suggests that proton therapy may offer improved dosimetry that translates to decreased post-operative morbidity in select patients with esophageal cancers.