Re-irradiation of lung cancer
Comparing rigid and deformable dose registration for high dose thoracic re-irradiation

https://doi.org/10.1016/j.radonc.2013.01.018Get rights and content

Abstract

For patients with locally-recurrent lung cancer, high dose thoracic re-irradiation can prolong survival. Deformable image registration improves the accuracy with which initial treatments are accounted for compared to rigid image registration. Using deformable image registration will improve correlative toxicity data, and may reduce toxicity for selected patients undergoing re-irradiation.

Section snippets

Material and methods

Between April 2003 and December 2011, all patients at the VU University Medical Center who received curative Re-RT, despite overlapping planning target volumes (PTV) were eligible for evaluation. No patients were excluded otherwise. Re-RT was only offered to patients after their case had been discussed and treatment recommended by, a thoracic multi-disciplinary tumor board. Medical ethics review was not sought as in The Netherlands, retrospective studies fall outside the scope of the Medical

Results

Ten consecutive patients with recurrent NSCLC underwent curative-intent Re-RT despite PTV overlap. The median patient age was 59 years (range 49–71). Patients were all of good performance status (WHO 0–1), and had undergone re-staging whole-body positron emission tomography and brain magnetic resonance imaging to confirm the absence of distant recurrence. The median time to Re-RT was 25 months (range 8–70). The median PTV overlap was 165 cc (range 2–470), representing a median 35% (range 1–89%) of

Discussion

Locally recurrent lung cancer is associated with extremely poor survival [9], [10]. Although the majority of clinicians consider Re-RT for local recurrence [11], this is seldom high dose and with curative intent [12]. Toxicity concerns are likely the primary reason for this. In this report, we quantified the spatial difference between RIR and DIR, finding differences are patient-specific and that for selected cases these can be significant. Additionally, we found DIR to be almost universally

Conflict of interest notification

The VU University Medical Center has a (non-reimbursed) research collaboration with Velocity Medical Solutions. The VU University Medical Center has received funds in its research collaboration with Varian Medical Systems. BJS and SUS have received honoraria and travel support from Varian Medical Systems. BJS has received honoraria and travel support from BrainLAB AG. SAS, GHG and JVSDK declare no personal conflicts of interest.

Acknowledgement

None.

References (16)

There are more references available in the full text version of this article.

Cited by (20)

  • Re-Irradiation of Recurrent Non-Small Cell Lung Cancer

    2021, Seminars in Radiation Oncology
    Citation Excerpt :

    In 3 studies, the initial radiotherapy technique was not specified.53-55 For re-irradiation, 7 studies used 3D conformal radiotherapy (3DCRT)25,29,34,50,54-56; 5 used 3DCRT plus intensity-modulated radiotherapy (IMRT), volumetric-modulated arc therapy (VMAT), or SABR27,32,37,45,57; 7 used IMPT5,22,35,36,39,46,53; and 16 used SABR.26,30,31,33,38,40-45,47-49,51,52 Eighty-eight percent of the studies using SABR as the initial treatment used repeat SABR for re-irradiation, with the remaining 12% using IMPT.

  • An improved method for analyzing and reporting patterns of in-field recurrence after stereotactic ablative radiotherapy in early-stage non-small cell lung cancer

    2020, Radiotherapy and Oncology
    Citation Excerpt :

    Other deformable registration packages are available and could be used, following careful implementation and validation and clear understanding of their performance and limitations. The advantages and pitfalls of various deformable image registration algorithms for lung cancer applications have been well described previously [18,29,31–34]. If Velocity AI or other commercial packages are not available in a given center, there are open source implementations of deformable image registration algorithms for lung cancer applications which are available as alternatives, which would have similar cautions on implementation and use, but which can provide similar potential to improve radiation treatment planning, delivery, and assessment and be used in approaches such as that reported here [35–37].

  • Evaluation of dose recalculation vs dose deformation in a commercial platform for deformable image registration with a computational phantom

    2018, Medical Dosimetry
    Citation Excerpt :

    It is a common practice in adaptive radiotherapy to collect cone beam computed tomography (CBCT) on a daily basis throughout the course of treatment. Because of daily variations, it is essential to register daily images with planning images and to deform contours and doses to monitor the actual dose administered to the organs.1-6 Whenever recurrent or re-treatment is required, it is particularly important, moreover, to map both doses and contours to a new set of planning images to evaluate the total dose administered to the patient.

  • Deformable image registration applied to lung SBRT: Usefulness and limitations

    2017, Physica Medica
    Citation Excerpt :

    Even if some attempts have been proposed, such as in [25] for lung, in [26] for brain or in [27] to include excision, to our knowledge, it is still an open problem. Besides those limitations, DIR has however been shown to be almost always more accurate than rigid registration [28,29]. Finally, it is also not clear what the consequences are of accumulating doses separated by several months but this is more of a radiobiology question than a DIR problem.

  • High-dose, conventionally fractionated thoracic reirradiation for lung tumors

    2014, Lung Cancer
    Citation Excerpt :

    An intensity modulated radiation therapy (IMRT) technique is now standard, either hybrid-IMRT [27] or volumetric modulated arc therapy and treatment is delivered using daily on-line image-guidance, with adaptive re-planning where needed. Currently, cumulative dose is estimated from original plans with the aid of rigid, and where appropriate deformable, co-registration [28]. To account for differences in dose and fractionation, equivalent doses in 2 Gray/fraction have been derived using the equation for biological effective dose (BED) with a α⧸β of 10 for tumor effects and the appropriate tissue-specific α⧸β ratio for late effects.

View all citing articles on Scopus
View full text