We demonstrate in a retrospective analysis of patients with MPM and treated at our institution with trimodality therapy that the use of postoperative highly conformal radiation techniques (HCRT) reduces local recurrence in comparison to 3DCRT. A recurrence analysis showed that in the case of 3DCRT 4 of 25 patients (16%) had a local recurrence in regions that were clearly underdosed according to current radiation protocols (doses ≥ 45 Gy are recommended, e.g. SAKK 17/04) in contrast to 0% of patients treated with HCRT. This supports the hypothesis that HCRT should improve local control in comparison to 3DCRT by improving target volume coverage. In our study patients treated with HCRT showed a tendency for improved progression free survival and local relapse free survival but did not benefit in terms of overall survival due to the high rates of distant relapses.
Local control is important in patients with MPM for symptom control, but also because some patients might benefit in terms of improved overall survival. Better local control after HCRT did not translate into improved overall survival in our patient series. Remarkably, the rate of death due to intercurrent disease, most often cardiac events, was higher after HCRT (29%) in comparison to 3DCRT (4%). Since cardiac sparing is rather improved with HCRT the most likely explanation for this difference is patient selection. The urgent research question, if postoperative radiotherapy impacts on overall survival after EPP, is addressed by a randomized study currently conducted in Switzerland, SAKK 1704. Patient accrual for this study was terminated in 2012 and the results are awaited.
Even after trimodality treatment local recurrence remains high in some patient series. In a retrospective series of 49 patients treated with 3D-conformal RT after EPP and chemotherapy 67% of all recurrences included the ipsilateral hemithorax and 25% of all recurrences were local only . Therefore improvement of radiotherapy is mandatory. In recent years radiotherapy has made enormous technical advances. More sophisticated highly conformal radiation techniques (HCRT) such as IMRT or rotational RT (VMAT) have become available and substituted for the older 3DCRT technique. The use of HCRT enables improvement in the dose distribution and target volume coverage. This is because with HCRT even complex target volumes such as the tumor bed of the costodiaphragmatic recess or the pericardium can be treated without or with little dose compromise and at the same time with optimal sparing of the normal tissue due to a steeper dose fall-off. Thus, the use of HCRT should intuitively improve treatment outcome in terms of local tumor control. Our data suggest indeed, that the use of HCRT bears considerable potential to improve on hemi-thoracic tumor control rates most likely due to improved target volume coverage.
The poor local control rates and high rates of in-field recurrences following 3DCRT in our cohort may be due to suboptimal dose coverage or the restriction of the target volume to avoid critical organs, both limitations inherent to the technique. After 3DCRT 4/24 (16.6%) in-field recurrences occurred in regions covered with only 30–43 Gy. In the case of 3DCRT mixed beams of photons and electrons were used to optimize dose coverage. The match of these beams often causes cold and hot spots of dose coverage. Poor matching during daily treatment can result in >20% dose inhomogeneity in the junction area . In addition, as the spinal cord is blocked when the tolerance dose of 45 Gy is reached, insufficient dose delivery to parts of the mediastinum has been observed, resulting in underdosage to the tumor bed .
Favorable tumor control after IMRT as part of a trimodality therapy has previously been reported by Rice et al. . The median overall survival of their 61 patients treated was 14.2 months with a locoregional recurrence rate of 13% and only 5% local in-field recurrences reported. The median dose prescribed was only 45 Gy, and half of all patients received doses even less than 45 Gy. The reason for the comparatively higher local control rate reported by Rice et al. in comparison to our study remains unclear. It may be explained by patient selection and the comparatively short median overall survival of 14.2 months in comparison to 20.8 months in the present series and by the retrospective study design. The shorter median overall survival reported by Rice et al. could be caused by more advanced tumor stages (40 T3, 8 T4, 26 N2), more aggressive subtypes (14 biphasic, 4 sarcomatoid) and the fact that neoadjuvant chemotherapy was not routinely administered.
With regard to toxicity the major dose limiting organ for postoperative radiotherapy of MPM is the contralateral lung. Lung complications such as radiation pneumonitis are likely to be higher with multi-field techniques such as IMRT or VMAT in comparison to 3DCRT, where opposed beams from 0 and 180 degrees are usually used, thereby optimally sparing the contralateral lung. With regard to dose escalation and lung sparing surgery, protons might prove superior to IMRT/VMAT. Severe complications of the lung with grade 4 and 5 pneumonitis after IMRT have been reported [7, 12]. Since then, special attention to the contralateral lung dose has been given during the treatment planning process and pneumonitis rates should be lower today. Intuitively, the use of HCRT should reduce toxicity and complication probabilities of esophagus, heart, liver and kidney, however no data with regard to these toxicity endpoints comparing both treatment techniques are available.
In recent years, the need for extensive surgery has been questioned, and less radical surgery has been advocated such as pleurectomy/decortication. In the context of reduced surgery, the anatomical situation makes it difficult for RT to be applied to the entire pleural space, however, it can still be considered as a targeted local postoperative option in case of incomplete resection. Future clinical studies are required to define the role of radiotherapy in combination with lung sparing surgery.