Towards spatial representations of dose distributions to predict risk of normal tissue morbidity after radiotherapy

Abstract

Design of clinical trials and definition of new standards of care in radiotherapy have historically been driven by the balance between doses delivered to the target volume against doses delivered to the healthy organs surrounding the tumour. State-of-the-art high-precision radiotherapy, together with other advances in cancer therapy, have led to improved clinical outcomes [[1]Bray F. Ferlay J. Soerjomataram I. Siegel R.L. Torre L.A. Jemal A. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries.CA Cancer J Clin. 2018; 68: 394-424https://doi.org/10.3322/caac.21492Crossref PubMed Scopus (51287) Google Scholar]. In the setting of conventionally fractionated radiotherapy, dose distributions within target volumes are reasonably uniform [[2]International Commission on Radiation Units and Measurements (ICRU). Prescribing, Recording, and Reporting Intensity-Modulated Photon-Beam Therapy (IMRT)(ICRU Report 83). https://icru.org/testing/reports/prescribing-recording-and-reporting-intensity-modulated-photon-beam-therapy-imrt-icru-report-83.Google Scholar]. In contrast, normal tissue dose distributions are highly non-uniform and their shape can be tailored while still achieving planning goals. These planning goals are typically formulated in terms of dose-volume constraints which drive optimization towards treatment plans deemed best for each patient. Visual inspection of how isodose lines are shaped around target volumes and into normal tissues is an integral part of plan checking and approval. Most radiotherapy plan optimization, evaluation and outcome analysis have been based on summarizing 3D dose distributions into dose-volume histograms (DVH). This approach has been successfully used to describe and utilize dose-volume-response relationships [[3]Olsson C.E. Jackson A. Deasy J.O. Thor M. A systematic post-QUANTEC review of tolerance doses for late toxicity after prostate cancer radiation therapy.Int J Radiat Oncol Biol Phys. 2018; 102: 1514-1532https://doi.org/10.1016/j.ijrobp.2018.08.015Abstract Full Text Full Text PDF PubMed Scopus (34) Google Scholar]. Limitations of DVH-based outcome assessments are well known. The DVH strips away geometry information, i.e. every region of normal tissue is seen as equally important, and no cross-talk between regions is considered. Clinical data support existence of regional effects, although explanations for these effects remain elusive [4Heemsbergen W.D. Incrocci L. Pos F.J. Heijmen B.J.M. Witte M.G. Local Dose effects for late gastrointestinal toxicity after hypofractionated and conventionally fractionated modern radiotherapy for prostate cancer in the HYPRO trial.Front Oncol. 2020; 10https://doi.org/10.3389/fonc.2020.00469Crossref Scopus (9) Google Scholar, 5Gagliardi G. Constine L.S. Moiseenko V. Correa C. Pierce L.J. Allen A.M. et al.Radiation dose-volume effects in the heart.Int J Radiat Oncol Biol Phys. 2010; 76: S77-S85https://doi.org/10.1016/j.ijrobp.2009.04.093Abstract Full Text Full Text PDF PubMed Scopus (481) Google Scholar]. Commonly used rationale to justify more sophisticated approaches relies on one or more of the following, and to a certain extent overlapping, arguments: 1) functional burden is not uniformly distributed in normal tissues; 2) stem cells capable of rescuing normal tissue function are primarily concentrated in particular parts of normal tissue; 3) organs which are contoured as a whole are in fact anatomically/geometrically sub-structured, e.g., lobes in lung or tracts in brain, and risk of morbidity does depend on which substructures receive dose; 4) parts of an organ are differentially sensitive to radiation; 5) morbidity is related to radiation-induced damage to organs other than ones for which DVHs are evaluated; 6) spatial distribution matters, e.g. small hot spots spread out on the rectal surface might have different consequences compared to larger, spatially clustered patterns. Dose-volume-response relationships that also incorporate the geometrical information of dose distributions have therefore received considerable interest. The first studies including geometrical information in normal tissue complication probability (NTCP) models utilized dose surface maps (DSMs), and were focused on hollow organs, in particular to study bladder and rectum dose-volume-response relationships following radiotherapy for prostate cancer [6Tucker S.L. Zhang M. Dong L. Mohan R. Kuban D. Thames H.D. Cluster model analysis of late rectal bleeding after IMRT of prostate cancer: a case-control study.Int J Radiat Oncol Biol Phys. 2006; 64: 1255-1264https://doi.org/10.1016/j.ijrobp.2005.10.029Abstract Full Text Full Text PDF PubMed Scopus (45) Google Scholar, 7Buettner F. Gulliford S.L. Webb S. Partridge M. Using dose-surface maps to predict radiation-induced rectal bleeding: a neural network approach.Phys Med Biol. 2009; 54: 5139-5153https://doi.org/10.1088/0031-9155/54/17/005Crossref PubMed Scopus (40) Google Scholar, 8Buettner F. Gulliford S.L. Webb S. Sydes M.R. Dearnaley D.P. Partridge M. Assessing correlations between the spatial distribution of the dose to the rectal wall and late rectal toxicity after prostate radiotherapy: an analysis of data from the MRC RT01 trial (ISRCTN 47772397).Phys Med Biol. 2009; 54: 6535-6548https://doi.org/10.1088/0031-9155/54/21/006Crossref PubMed Scopus (66) Google Scholar]. Therefore, utilization of DSMs has helped to demonstrate that the dose delivered to the caudal part together with shape and extension of the hot spots over rectal wall are related to an exacerbation of morbidity [9Shelley L.E.A. Scaife J.E. Romanchikova M. Harrison K. Forman J.R. Bates A.M. et al.Delivered dose can be a better predictor of rectal toxicity than planned dose in prostate radiotherapy.Radiother Oncol. 2017; 123: 466-471https://doi.org/10.1016/j.radonc.2017.04.008Abstract Full Text Full Text PDF PubMed Scopus (36) Google Scholar, 10Shelley L.E.A. Sutcliffe M.P.F. Thomas S.J. Noble D.J. Romanchikova M. Harrison K. et al.Associations between voxel-level accumulated dose and rectal toxicity in prostate radiotherapy.Phys Imag Radiat Oncol. 2020; 14: 87-94https://doi.org/10.1016/j.phro.2020.05.006Abstract Full Text Full Text PDF PubMed Scopus (9) Google Scholar, 11Wortel R.C. Witte M.G. van der Heide U.A. Pos F.J. Lebesque J.V. van Herk M. et al.Dose-surface maps identifying local dose-effects for acute gastrointestinal toxicity after radiotherapy for prostate cancer.Radiother Oncol. 2015; 117: 515-520https://doi.org/10.1016/j.radonc.2015.10.020Abstract Full Text Full Text PDF PubMed Scopus (52) Google Scholar, 12Casares-Magaz O. Bülow S. Pettersson N.J. Moiseenko V. Pedersen J. Thor M. et al.High accumulated doses to the inferior rectum are associated with late gastro-intestinal toxicity in a case-control study of prostate cancer patients treated with radiotherapy.Acta Oncol. 2019; 58: 1543-1546https://doi.org/10.1080/0284186X.2019.1632476Crossref PubMed Scopus (10) Google Scholar, 13Onjukka E. Fiorino C. Cicchetti A. Palorini F. Improta I. Gagliardi G. et al.Patterns in ano-rectal dose maps and the risk of late toxicity after prostate IMRT.Acta Oncol. 2019; 58: 1757-1764https://doi.org/10.1080/0284186X.2019.1635267Crossref PubMed Scopus (11) Google Scholar]. Similarly, doses delivered to the trigone area and the urethra were found to play a major role in the overall worsening of genito-urinary symptoms following radiotherapy for prostate cancer [14Palorini F. Cozzarini C. Gianolini S. Botti A. Carillo V. Iotti C. et al.First application of a pixel-wise analysis on bladder dose-surface maps in prostate cancer radiotherapy.Radiother Oncol. 2016; 119: 123-128https://doi.org/10.1016/j.radonc.2016.02.025Abstract Full Text Full Text PDF PubMed Scopus (42) Google Scholar, 15Mylona E. Acosta O. Lizee T. Lafond C. Crehange G. Magné N. et al.Voxel-based analysis for identification of urethrovesical subregions predicting urinary toxicity after prostate cancer radiation therapy.Int J Radiat Oncol Biol Phys. 2019; 104: 343-354https://doi.org/10.1016/j.ijrobp.2019.01.088Abstract Full Text Full Text PDF PubMed Scopus (27) Google Scholar, 16Yahya N. Ebert M.A. House M.J. Kennedy A. Matthews J. Joseph D.J. et al.Modeling urinary dysfunction after external beam radiation therapy of the prostate using bladder dose-surface maps: evidence of spatially variable response of the bladder surface.Int J Radiat Oncol Biol Phys. 2017; 97: 420-426https://doi.org/10.1016/j.ijrobp.2016.10.024Abstract Full Text Full Text PDF PubMed Scopus (37) Google Scholar, 17Ghadjar P. Zelefsky M.J. Spratt D.E. Munck P. af Rosenschöld Oh J.H. Hunt M. et al.Impact of dose to the bladder trigone on long-term urinary function after high-dose intensity modulated radiation therapy for localized prostate cancer.Int J Radiat Oncol Biol Phys. 2014; 88: 339-344https://doi.org/10.1016/j.ijrobp.2013.10.042Abstract Full Text Full Text PDF PubMed Scopus (107) Google Scholar, 18Heemsbergen W.D. Al-Mamgani A. Witte M.G. van Herk M. Pos F.J. Lebesque J.V. Urinary obstruction in prostate cancer patients from the Dutch trial (68 Gy vs. 78 Gy): relationships with local dose, acute effects, and baseline characteristics.Int J Radiat Oncol Biol Phys. 2010; 78: 19-25https://doi.org/10.1016/j.ijrobp.2009.07.1680Abstract Full Text Full Text PDF PubMed Scopus (91) Google Scholar]. Overall, DSMs are often used in an empirical approach, exploiting significant differences in dose distribution patterns between patients with and without morbidity to identify regions which might correspond to substructures governing the dose response. In the search for methods to include geometrical information into NTCP models, a more theoretical or deterministic approach can also be considered. A deterministic approach hypothesizes that an organ/tissue is composed by subunits with different radiosensitivity, with one (or more) playing a major role in the manifestation of morbidity. Therefore, damage to an identified substructure(s) may trigger manifestation of the observed morbidity. This issue of the journal presents the first study describing the impact of dose to heart substructures on the overall survival rate [[19]McWilliam A. Dootson C. Graham L. Banfill K. Abravan A. van Herk M. Dose surface maps of the heart can identify regions associated with worse survival for lung cancer patients treated with radiotherapy.Phys Imag Radiat Oncol. 2020; 15: 46-51https://doi.org/10.1016/j.phro.2020.07.002Abstract Full Text Full Text PDF Scopus (12) Google Scholar]. Cardiac disease associated with radiation has been observed in cancer patients treated to thoracic and abdominal regions, i.e. breast cancer and lymphoma patients, because they are typically long-term survivors, providing most of the outcome data. The QUANTEC report on the heart provided dose-volume guidelines for two endpoints: pericarditis and long-term cardiac mortality [[5]Gagliardi G. Constine L.S. Moiseenko V. Correa C. Pierce L.J. Allen A.M. et al.Radiation dose-volume effects in the heart.Int J Radiat Oncol Biol Phys. 2010; 76: S77-S85https://doi.org/10.1016/j.ijrobp.2009.04.093Abstract Full Text Full Text PDF PubMed Scopus (481) Google Scholar]. For the former, mean dose < 26 Gy and V30Gy < 46% were suggested. Cardiac mortality, in contrast to pericarditis, appears to exhibit a weak volume effect. This observation has led to substantial changes in how radiotherapy is delivered to left-sided breast cancer patients. Heart blocks including the “no heart in beams-eye-view” policies have been adopted, deep inspiration breath hold and treating prone techniques have also been used to spare the heart. The QUANTEC heart paper proposed a V25Gy < 10% constraint to control the “tail” of the DVH [[5]Gagliardi G. Constine L.S. Moiseenko V. Correa C. Pierce L.J. Allen A.M. et al.Radiation dose-volume effects in the heart.Int J Radiat Oncol Biol Phys. 2010; 76: S77-S85https://doi.org/10.1016/j.ijrobp.2009.04.093Abstract Full Text Full Text PDF PubMed Scopus (481) Google Scholar], however, this constraint has been later shown to not fully protect against the probability of cardiac mortality < 1% in breast cancer patients [[20]Moiseenko V. Einck J. Murphy J. Ödén J. Bjöhle J. Uzan J. et al.Clinical evaluation of QUANTEC guidelines to predict the risk of cardiac mortality in breast cancer patients.Acta Oncol. 2016; 55: 1506-1510https://doi.org/10.1080/0284186X.2016.1234067Crossref PubMed Scopus (11) Google Scholar]. This also underlines another deficiency in using DVH constraints. Various endpoints exhibit different dose-volume response. Consequently optimizing, e.g. mean dose, will reduce probability of pericarditis but is not sufficient to control the risk of cardiac mortality. Identifying an endpoint of the greatest concern or with the tightest constraints is one possible way but it may become too restrictive. For the heart also the initial definition and delineation of the organ at risk is challenging. Specifically, the whole heart can be contoured, while also the pericardium (often defined as a “shell” expansion of the heart contour), the left ventricle, coronary vessels or specifically the left anterior descending coronary artery are used in different protocols and institutions [[5]Gagliardi G. Constine L.S. Moiseenko V. Correa C. Pierce L.J. Allen A.M. et al.Radiation dose-volume effects in the heart.Int J Radiat Oncol Biol Phys. 2010; 76: S77-S85https://doi.org/10.1016/j.ijrobp.2009.04.093Abstract Full Text Full Text PDF PubMed Scopus (481) Google Scholar]. Different endpoints have been connected to anatomical features and functionality, for example the pericardium is likely appropriate for pericardial effusion and the left anterior descending artery for ischemic heart disease [[20]Moiseenko V. Einck J. Murphy J. Ödén J. Bjöhle J. Uzan J. et al.Clinical evaluation of QUANTEC guidelines to predict the risk of cardiac mortality in breast cancer patients.Acta Oncol. 2016; 55: 1506-1510https://doi.org/10.1080/0284186X.2016.1234067Crossref PubMed Scopus (11) Google Scholar]. However, dose-volume parameters for different structures behave in a correlated manner which complicates search for the “guilty” party, and a full mechanistic understanding of how morbidity develops is lacking. Searching for regional effects may provide us with a guide to selective sparing of either structures, or geometrically defined sections of the organ. Because multiple anatomically/geometrically defined structures/regions must be analysed for different endpoints, automation might be required. Auto-contouring not only allows us to tackle a laborious task of delineating structures which are not routinely delineated in clinical practice, it may also improve consistency [[21]Haq R. Hotca A. Apte A. Rimner A. Deasy J.O. Thor M. Cardio-pulmonary substructure segmentation of radiotherapy computed tomography images using convolutional neural networks for clinical outcomes analysis.Phys Imag Radiat Oncol. 2020; 14: 61-66https://doi.org/10.1016/j.phro.2020.05.009Abstract Full Text Full Text PDF Scopus (23) Google Scholar]. Additionally, auto-contouring tools included in retrospective analysis will pave the way to interrogating outcomes data and will further promote prospective use for selective regional sparing [[22]Jung J.W. Lee C. Mosher E.G. Mille M.M. Yeom Y.S. Jones E.C. et al.Automatic segmentation of cardiac structures for breast cancer radiotherapy.Phys Imag Radiat Oncol. 2019; 12: 44-48https://doi.org/10.1016/j.phro.2019.11.007Abstract Full Text Full Text PDF Scopus (13) Google Scholar]. An overall aim of dose-surface based outcome studies is to identify the substructures of the organ that play a key role in dose–response relationships, and from there provide robust metrics to be used in plan optimisation. Most studies in this field have a analysis to significant dose differences L.E.A. Sutcliffe M.P.F. Thomas S.J. Noble D.J. Romanchikova M. Harrison K. et al.Associations between voxel-level accumulated dose and rectal toxicity in prostate radiotherapy.Phys Imag Radiat Oncol. 2020; 14: 87-94https://doi.org/10.1016/j.phro.2020.05.006Abstract Full Text Full Text PDF PubMed Scopus (9) Google Scholar, 11Wortel R.C. Witte M.G. van der Heide U.A. Pos F.J. Lebesque J.V. van Herk M. et al.Dose-surface maps identifying local dose-effects for acute gastrointestinal toxicity after radiotherapy for prostate cancer.Radiother Oncol. 2015; 117: 515-520https://doi.org/10.1016/j.radonc.2015.10.020Abstract Full Text Full Text PDF PubMed Scopus (52) Google Scholar, 12Casares-Magaz O. Bülow S. Pettersson N.J. Moiseenko V. Pedersen J. Thor M. et al.High accumulated doses to the inferior rectum are associated with late gastro-intestinal toxicity in a case-control study of prostate cancer patients treated with radiotherapy.Acta Oncol. 2019; 58: 1543-1546https://doi.org/10.1080/0284186X.2019.1632476Crossref PubMed Scopus (10) Google Scholar, 13Onjukka E. Fiorino C. Cicchetti A. Palorini F. Improta I. Gagliardi G. et al.Patterns in ano-rectal dose maps and the risk of late toxicity after prostate IMRT.Acta Oncol. 2019; 58: 1757-1764https://doi.org/10.1080/0284186X.2019.1635267Crossref PubMed Scopus (11) Google Scholar, F. Cozzarini C. Gianolini S. Botti A. Carillo V. Iotti C. et al.First application of a pixel-wise analysis on bladder dose-surface maps in prostate cancer radiotherapy.Radiother Oncol. 2016; 119: 123-128https://doi.org/10.1016/j.radonc.2016.02.025Abstract Full Text Full Text PDF PubMed Scopus (42) Google Scholar, 15Mylona E. Acosta O. Lizee T. Lafond C. Crehange G. Magné N. et al.Voxel-based analysis for identification of urethrovesical subregions predicting urinary toxicity after prostate cancer radiation therapy.Int J Radiat Oncol Biol Phys. 2019; 104: 343-354https://doi.org/10.1016/j.ijrobp.2019.01.088Abstract Full Text Full Text PDF PubMed Scopus (27) Google Scholar]. data of radiotherapy doses relies on the of delivered dose over of patients with and without the morbidity endpoint of interest. a of which is based on the of the organ and the of the endpoint of interest. For hollow organs or are for an anatomical of In are to the effects, more sophisticated methods based on techniques may E. Acosta O. Lizee T. Lafond C. Crehange G. Magné N. et al.Voxel-based analysis for identification of urethrovesical subregions predicting urinary toxicity after prostate cancer radiation therapy.Int J Radiat Oncol Biol Phys. 2019; 104: 343-354https://doi.org/10.1016/j.ijrobp.2019.01.088Abstract Full Text Full Text PDF PubMed Scopus (27) Google Scholar]. these typically have of be to A is the of the delivered dose to tissues in the of often only treatment planning and dose distributions are which in the of a highly organ may only a treatment computed tomography or may improve these estimates and the between doses and the morbidity endpoints [9Shelley L.E.A. Scaife J.E. Romanchikova M. Harrison K. Forman J.R. Bates A.M. et al.Delivered dose can be a better predictor of rectal toxicity than planned dose in prostate radiotherapy.Radiother Oncol. 2017; 123: 466-471https://doi.org/10.1016/j.radonc.2017.04.008Abstract Full Text Full Text PDF PubMed Scopus (36) Google Scholar, 12Casares-Magaz O. Bülow S. Pettersson N.J. Moiseenko V. Pedersen J. Thor M. et al.High accumulated doses to the inferior rectum are associated with late gastro-intestinal toxicity in a case-control study of prostate cancer patients treated with radiotherapy.Acta Oncol. 2019; 58: 1543-1546https://doi.org/10.1080/0284186X.2019.1632476Crossref PubMed Scopus (10) Google Scholar, M. L. C. et of rectum and bladder morbidity following radiotherapy of prostate cancer based on dose Oncol. Full Text Full Text PDF PubMed Scopus Google Scholar, S. Thor M. Rimner A. N. Zhang L. et of accumulated dose and associated anatomical changes of using radiotherapy of lung Imag Radiat Oncol. 2020; Full Text Full Text PDF PubMed Scopus Google Scholar]. of the of an observed in doses over the outcome the multiple by the of is using a approach, patients between the morbidity and C. Witte M.G. W.D. van Herk M. for based data in Oncol. PubMed Scopus Google Scholar]. The associated significant dose a of the local dose by and in the observation which this be that using these be with for a of which have a dose by to but significant and survival can be to the observed dose patterns in terms of an anatomical or a and likely the most of the dose data one a of prostate cancer patients been with an between control probability and risk of and one rectum morbidity dose with have received dose to a of the rectum to morbidity and dose distributions might compared to from patients with more These dose regions the fact that dose and the dose to and of the external beam radiotherapy R.C. Witte M.G. van der Heide U.A. Pos F.J. Lebesque J.V. van Herk M. et al.Dose-surface maps identifying local dose-effects for acute gastrointestinal toxicity after radiotherapy for prostate cancer.Radiother Oncol. 2015; 117: 515-520https://doi.org/10.1016/j.radonc.2015.10.020Abstract Full Text Full Text PDF PubMed Scopus (52) Google Scholar, F. Cozzarini C. Gianolini S. Botti A. Carillo V. Iotti C. et al.First application of a pixel-wise analysis on bladder dose-surface maps in prostate cancer radiotherapy.Radiother Oncol. 2016; 119: 123-128https://doi.org/10.1016/j.radonc.2016.02.025Abstract Full Text Full Text PDF PubMed Scopus (42) Google Scholar, 15Mylona E. Acosta O. Lizee T. Lafond C. Crehange G. Magné N. et al.Voxel-based analysis for identification of urethrovesical subregions predicting urinary toxicity after prostate cancer radiation therapy.Int J Radiat Oncol Biol Phys. 2019; 104: 343-354https://doi.org/10.1016/j.ijrobp.2019.01.088Abstract Full Text Full Text PDF PubMed Scopus (27) Google Scholar]. of these as dose regions, e.g. these patterns into a treatment planning to dose the treatment plans and to clinical observed patterns as the for an of the the clinical manifestation of e.g. in the way these patterns with substructures of the organ. prospective be to these into NTCP models which may improve radiotherapy for patients. be to the possible impact of differences between planned and delivered driven by organ In significant local dose effects are found in the of organs is more [14Palorini F. Cozzarini C. Gianolini S. Botti A. Carillo V. Iotti C. et al.First application of a pixel-wise analysis on bladder dose-surface maps in prostate cancer radiotherapy.Radiother Oncol. 2016; 119: 123-128https://doi.org/10.1016/j.radonc.2016.02.025Abstract Full Text Full Text PDF PubMed Scopus (42) Google Scholar, F. Botti A. Carillo V. Gianolini S. Improta I. Iotti C. et maps and urinary with to in local dose 2016; Full Text Full Text PDF PubMed Scopus (23) Google Scholar]. This the of significant differences in the doses between patients with and without morbidity because in more the planned dose is to the delivered An is by the of the bladder to variable with the while the This fact systematic vs. delivered in the dose region which to the bladder F. Botti A. Carillo V. Gianolini S. Improta I. Iotti C. et maps and urinary with to in local dose 2016; Full Text Full Text PDF PubMed Scopus (23) Google Scholar]. on local dose effects for the bladder also be to the of the features at the in the with doses to the bladder out as not significant to the between planned and delivered studies on organ and on differences between planned and delivered dose in between local dose effects and effects driven by the of delivered doses in regions A. M. G. Acosta O. Lafond C. P. et of the and in the in J Radiat Oncol. 2014; Full Text Full Text PDF Google Scholar, O. Moiseenko V. A. Pettersson N.J. Thor M. R. et al.Associations between volume changes and spatial dose metrics for the urinary bladder local for prostate Oncol. 2017; PubMed Scopus Google Scholar]. DSMs might of normal tissue dose response from data be by studies on organ on of differences between planned and delivered dose and on the possible role of the regions identified as associated with an risk of radiation-induced morbidity. on a of will pave the way to clinical of the The that they have no or relationships that have to the in this

Topics & Concepts