Research ArticleClinical Studies

Impact of Anatomical Position Errors on Dose Distribution in Head and Neck Radiotherapy and Robust Image Registration Against Anatomical Changes

KOUTA HIROTAKI, KENTO TOMIZAWA, SHUNSUKE MORIYA, MASASHI ITO and TAKEJI SAKAE
Anticancer Research April 2023, 43 (4) 1827-1834; DOI: https://doi.org/10.21873/anticanres.16336

Abstract

Background/Aim: This study pursued two goals: Firstly, to search for anatomical structures strongly correlating with dose deterioration, and secondly to investigate the effectiveness of image registration focusing on critical anatomy by comparing it with a conventional method. The aim was to achieve robust image registration to correct for anatomical changes during treatment. Patients and Methods: Twenty patients with head and neck cancer were enrolled, and 68 simulation computed tomography (CT) and rescan CT image sets were retrospectively analyzed. Forty volumetric-modulated arc therapy and intensity-modulated proton therapy plans were generated and recalculated according to the rescan CT to evaluate the dose effects of anatomical changes. Correlation coefficients were calculated for the relationships between the six-axis motion of the anatomy and the dose indices for the clinical target volume (CTV) and organs at risk. In the image registration, we compared a conventional method and target-based registration that limited the registration range to the CTV and vertebrae. Results: The CTV coverage and spinal cord dose were correlated with the position error associated with the pitch and vertical position of the vertebrae, and the parotid gland and oral cavity dose were strongly correlated with the position error associated with the roll of the clivus and mandible. The target registration improved CTV coverage and suppressed the increase in dose to organs at risk compared with conventional methods. Conclusion: Monitoring vertebral alignment, the assessment and correction of positioning errors associated with the clivus and mandible position errors are important to ensure the quality of daily treatment. Target-based registration may allow for more robust image registration.

Key Words:

In the case of head and neck cancer, intensity-modulated proton therapy (IMPT) can achieve better dose reduction to normal tissues than intensity-modulated radiotherapy (IMRT) using X-ray beams (1-3). However, when IMPT and IMRT are performed on the head and neck region, the dose distribution deteriorates because of anatomical changes and setup errors occurring over the treatment period (4-6). Moreover, these anatomical changes have a greater impact on dose delivery in proton beam therapy compared with photon beam therapy (7-11). Robust planning is therefore employed in IMPT to compensate for the effects of anatomical variations and setup errors (12-16). However, although the plan so produced may be robust in countering anatomical variations and positioning errors, the dose to organs at risk (OARs) is generally increased, and the plan cannot accommodate large changes beyond the compensation range. Adaptive radiation therapy (ART), which uses a rescan computed tomography (CT) acquired during the treatment period to modify the plan, has also been employed to update the treatment plan for the effects of anatomical variations (17, 18). Several studies using ART have been reported, and various strategies have been devised (19). One of the typical ART strategies is the online adaptive strategy, in which the dose distribution is updated in daily treatment and provides a more robust and superior dose distribution (18, 20, 21). However, online adaptive ART requires high-quality imaging in the treatment room, creation of the clinical target volume (CTV) and OAR contours during treatment, recalculation of the particle transport and optimization, dose distribution verification, and physician approval. This requires advanced technology and is resource-intense. In contrast, image-guided radiation therapy (IGRT) and patient repositioning are simpler adaptive treatments accounting for anatomical changes, which do not require complicated procedures and can be applied in most clinics. However, because IGRT and repositioning are performed on the basis of the therapist’s experience, the quality of the procedures depends on the operator. In this study, we searched for anatomical structures showing strong associations with deterioration in the dose distribution to maximize the improvement of the dose distribution through IGRT and repositioning. Furthermore, we performed image registration using regions of interest (ROIs) fitted to highly relevant anatomical structures (target-based registration) and evaluated the effectiveness of this target-based registration by comparing it with a conventional method.

Patients and Methods

Patients and imaging datasets. Twenty patients with head and neck cancer were enrolled (Table I), and 20 simulation CT and 48 rescan CT images were retrospectively analyzed. The patients were treated with IMRT or IMPT, and CT images were acquired 2-4 times during the treatment period. All patients were immobilized using a four-point thermoplastic mask covering the shoulder and a patient-specific pillow. CT datasets were acquired using an Aquilion One scanner (Canon Medical Systems, Tochigi, Japan). Rescan CT images of six patients were acquired once at the 20th treatment, and rescan CT images of 14 patients were acquired at the 10th, 20th, and 30th treatments. Patient information that could identify an individual was anonymized. Although patient-informed consent was not required, details of this study were published on the center’s homepage and patients were given the opportunity to refuse to participate in the study. The contents of this study, including the investigation procedure and the handling of patient information, were approved by the Institutional Review Board of the National Cancer Center Hospital East (IRB No. 2020-282).

View this table:
Table I.

Patient characteristics.

Treatment planning. Forty volume modulated arc therapy (VMAT) and IMPT plans based on the initial CT images of the 20 patients were created. VMAT plans were created using two arcs, and the IMPT plans were created using 2-4 fields. All treatments were planned using an Eclipse treatment planning system (Varian Medical Systems, Palo Alto, CA, USA) by a well-trained therapist and medical physicist. The CTV and OARs were contoured on the simulation CT images by a radiation oncologist, and the CTV and OARs on the rescan CT were defined by the medical physicist using a deformed ROI based on the contouring by the radiation oncologist. All treatment plans were generated with 35 fractions of 2 Gy [relative biological effectiveness (RBE)] to give a total dose of 70 GyRBE. IMPT plans were generated to include a robustness to range uncertainty of 3.5% and isocenter uncertainty of 3 mm, using multiple field optimization. All plans were designed in accordance with the clinical protocol: CTV (high-risk), >98% of the volume to receive >95% of 70 GyRBE; CTV (low-risk), >98% of the volume to receive >95% of 54 GyRBE; CTV, ≤2% to receive <110% of 70 GyRBE; maximum dose to spinal cord <45 GyRBE; maximum dose to brain stem <54 GyRBE; mean dose to parotid gland <26 GyRBE; dose to cochlea <40 GyRBE; mean dose to oral cavity <35 GyRBE; ≤2% of lower jawbone to receive <105% of 70 GyRBE.

Anatomical structures and corrections. The typical anatomical structures of the head and neck region were selected for analysis: The clivus, second cervical vertebra, fifth cervical vertebra, first thoracic vertebra, lower jawbone, sternum, hyoid bone, and both shoulders. The positioning errors were obtained by calculating the differences in the six-axis vectors [anterior–posterior (A-P), superior–inferior (S-I), left–right (L-R), yaw, pitch, and roll] for each anatomical structure between the initial and replanning CT images. The positioning error determined according to the imaging positions in both CT images was corrected by triaxial translation. Calculation of the position error was performed by placing a rectangular ROI on the anatomical structure and using rigid image registration to align the ROIs in the two image sets. MIM Maestro® software (MIM Software Inc., Cleveland, OH, USA) was used for the rigid image registration algorithm. The three-dimensional vector of the position error for translational displacements (A-P, S-I, R-L) was calculated using Equation 1 below.

Embedded Image (Eq. 1)

Correlation coefficients between anatomical structures and dose indices. Pearson’s correlation coefficients were calculated using a combination of the eight anatomical structures and their respective four correction factors (3D vector, yaw, pitch and roll) and the 12 dose indices (CTVhigh D98, CTVlow D98, CTVhigh D2, spinal cord D1cc, parotid gland D50, oral cavity D50, brain stem D1cc, lower jawbone D2, larynx D50, conformity index, maximum dose, minimum dose in CTV). Differences in dose index were calculated for VMAT and IMPT by subtracting the respective recalculated plans based on the rescan CT from the initial plans. Differences in the corrected amounts and dose indices were calculated for 48 sets of initial CT and rescan CT, and correlation coefficients between each anatomical structure and important dose indices on the dose distribution were obtained.

Automatic image registration. Automatic registration was performed using the box image registration function of MIM Maestro®, and was used to calculate the dose distribution after the six-axis correction performed at the treatment couch. The target ROIs for automatic registration were set as follows: i) The entire patient image (all-image registration); ii) limited to the range covering the PTV (PTV registration), iii). limited to the clivus-C6 range (spine registration). Automatic registration was performed by transforming the rescan CT to the initial planning CT using each target ROI, and the CT after six-axis correction was reconstructed as a new CT using MIM Maestro®. The slice thickness of the new reconstructed CT images was 2 mm, which was the same as the initial planning CT. The new reconstructed CT images were transferred to the Eclipse treatment planning system and the treatment plans were recalculated using the initial VMAT and IMPT plans. The dose distributions obtained from the three automatic registration methods (six degrees of freedom) and three-axis correction were compared with the initial plan dose index.

Statistical analysis. Pearson’s correlation coefficient was considered weak, intermediate, and strong at <0.4, 0.4-0.7, and >0.7, respectively. The average dose differences from the initial plan of VMAT or IMPT were assessed using pairwise comparisons with t-tests. Bonferroni corrections were applied for multiple comparisons. All the statistical analyses were two-sided and were performed using R v. 4.2.0; statistical significance was set at p<0.05.

Results

Position error of the anatomical structures according to elapsed treatment days. Figure 1 and Supplementary Table I show the position error of each of the anatomical structures on CT images acquired at 10, 20, and 30 fractions. In the R-L direction, the position error was small, and was not dependent on the treatment day. In the S-I direction, the positional errors of the hyoid bone and shoulder were relatively large. In the A-P direction, the vertebrae (C2, C5, TH1), sternum, and shoulder position errors were large, and those of the C2, C5, TH1 and sternum increased in association with elapsed treatment days from initial treatment. In the yaw and roll directions, position error was random, and was not dependent on the treatment day. In the pitch direction, the errors for C2, TH1, hyoid bone and sternum increased with the elapsed treatment days. especially those for C2, for which the average position error exceeded 2.5 degrees at 30 fractions.

Figure 1.
Figure 1.

The average position error of anatomical structures in translation and rotation from simulation computed tomography for 20 patients. In the vertical plane and pitch, increasing position error of spinal structures was seen with elapsed treatment.

Correlations between dose index and anatomical position errors. Supplementary Table II and Supplementary Table III show the correlation coefficients between anatomical position errors and dose indices for VMAT and IMPT planning. For the CTV of high-risk coverage, the correlation coefficients of C2-pitch, C5-3Dvector, and TH1-3Dvector were −0.46, −0.43, and −0.43, respectively for VMAT, and – 0.39, −0.41, and −0.35 for IMPT, indicating intermediate correlations. These anatomical indices also negatively correlated with the minimum CTV dose. The correlation coefficient between hyoid bone-pitch and the CTV for high-risk D98 was −0.44 for VMAT, while the correlation coefficients between hyoid bone-3D vector and the conformity index were 0.47 for VMAT and 0.38 for IMPT. For the OAR dose, the correlation coefficients between parotid gland D50 dose and C2-roll, lower jawbone-roll, and clivus-roll were 0.55, 0.74, and 0.74, respectively, for VMAT, and 0.72, 0.71, and 0.81 for IMPT, indicating strong correlations. The correlation coefficients between oral cavity dose D50 and C2-roll, lower jawbone-roll, and clivus-roll were 0.74, 0.53, and 0.65, respectively, for VMAT, but no strong correlation was found for IMPT. The correlation coefficients between spinal cord D1cc and the C5-3Dvector and TH1-3Dvector were 0.39 and 0.45, respectively, for VMAT, but there were weak correlations for IMPT. For the hyoid bone, sternum, and shoulder, weak correlations with dose indices was found for either VMAT or IMPT.

Image registration strategy and dose index. We evaluated three-axis image registration and three methods for six-axis automatic image registration with different registration ROIs. Figure 2 shows boxplots of the CTV high-risk D98, CTV high-risk D2, CTV low-risk D98, spinal cord D1cc, parotid D50, and oral cavity D50 in VMAT. In the CTV high-risk D2, the average increase was more than 1% with the three-axis correction, but less than 0.5% with all three methods for six-axis correction. The coverage of CTV high-risk D98 improved slightly with six-axis correction compared with three-axis correction. For spinal cord-D1cc, the average dose increased by 2 Gy, and the maximum average dose increased by more than 12 Gy with three-axis correction. Six-axis correction with spine registration showed the best performance in respect to the mean and standard deviation of the dose. For parotid D50, an average dose increase of 2 Gy and a maximum average dose of more than 25 Gy was found with the three-axis correction. Of the three six-axis methods, the six-axis correction with spine registration provided the lowest increase in the mean parotid dose. Figure 3 shows boxplots of CTV high-risk D98, CTV high-risk D2, CTV low-risk D98, spinal cord D1cc, parotid D50, and oral cavity D50 in IMPT. Coverage of the CTV high-risk D98 improved slightly with six-axis correction compared with three-axis correction. However, the other dose indices showed no significant differences between the six-axis correction and three-axis correction.

Figure 2.
Figure 2.

The average dose difference from the initial plan in volumetric-modulated arc therapy. Boxplots represent each image registration method focusing on different regions of interest (ROIs). Spine: Translation and rotation correction focusing on spinal ROIs, 6 degrees of freedom (DoF); PTV: 6 DoF focusing on planning target volume ROI; All: 6 DoF using all image information (without focusing on ROIs); 3DoF: translation correction only. The boxes represent the interquartile range and the horizontal lines the median values, crosses indicate the average values. Whisker lines include data within the 1.5-fold of the interquartile range; dots are outliers. CTV: Clinical target volume; D2/50/98: dose to 2%/50%/98% of the volume; D1cc: Dose delivered to 1 cm3 volume. *Significantly different.

Figure 3.
Figure 3.

The average dose difference from the initial plan in intensity-modulated proton therapy. Boxplots represent each image registration method focusing on different regions of interest (ROIs). Spine: Translation and rotation correction focusing on spinal ROIs, 6 degrees of freedom (DoF); PTV: 6 DoF focusing on planning target volume ROI; All: 6 DoF using all image information (without focusing on ROIs); 3DoF: translation correction only. The boxes represent the interquartile range and the horizontal lines the median values, crosses indicate the average values. Whisker lines include data within the 1.5-fold of the interquartile range, dots are outliers. CTV: Clinical target volume; D2/50/98: dose to 2%/50%/98% of the volume; D1cc: Dose delivered to 1 cm3 volume.

Discussion

In this study, setup errors and their effects on dose distribution were retrospectively investigated in 20 patients with head and neck cancer. The relationship between setup error and dose deterioration was previously investigated in patients with nasopharyngeal cancer (22), but the study did not include rotational motion as a setup error, and position error was evaluated only at the isocenter position. In contrast, we investigated the relationship between the six-axis movement of each anatomical structure and the dose distribution. In addition, we evaluated the effectiveness of target registration focusing on identified anatomical structures. In the relationship between position error and treatment day, the position errors of C2-pitch and C2, C5, and TH1 in the vertical direction increased with the elapsed treatment days. We believe this may be because the thickness of the posterior neck and back decreased due to weight loss during the treatment period, and the vertebrae were therefore displaced downward. C2-pitch, C5-3D vector, and TH1-3D vector were associated with decreased CTV coverage. Therefore, the positional errors of C2-pitch, C5-3D vector, and TH1-3D vector increased because of the decrease in thickness of the posterior neck and back, and changes in these anatomical structures may be factors that can predict a decrease in CTV coverage. Spine registration with an emphasis on the vertebrae improved the CTV coverage. We believe this was because the spine registration corrected the downward displacement of the vertebrae and brought the vertebral alignment at treatment closer to that of the initial CT. The C5-3D vector and TH1-3D vector correlated with the dose increase to the spinal cord, and the spine registration suppressed the dose increase to the spinal cord the most out of the three methods evaluated. The increase in spinal cord dose is caused by a shift of the high-dose area towards the vertebrae, which in turn is caused by displacement of the vertebral alignment due to the reduction in thickness of the posterior neck and back. We believe that the spine registration corrected the downward displacement of the vertebrae well, and thereby suppressed the increase in the spinal cord dose.

An increase in parotid gland dose strongly correlated with C2-roll, lower jawbone-roll, and clivus-roll. A high correlation between mandibular rotation and parotid gland dose is consistent with the report by Otsuka et al. (23). However, the position errors in the roll direction were less related to the elapsed treatment time. Position errors of the C2, mandible, and clivus in the roll direction mean that the orientation of the patient’s face was different from that at the time of the initial planning. Because no correlation with the elapsed treatment time was observed, these errors appear to have been caused by daily patient positioning. Change in a patient’s position resulting in rotation of the face with the body facing forward cannot be corrected by six-axis correction (24). Spine registration was able to minimize the increase in the mean dose to the parotid gland; since the image registration range was limited to the vertebrae (C1-6) in the spine registration, anatomical structures such as the clivus and C2 were emphasized in the six-axis correction, and we believe that the orientation of the patient’s face was therefore well corrected. In the VMAT plans, the spine registration and CTV registration (six-axis correction using limited ROI) contributed to improved CTV coverage, fewer hot-spots, and lower OAR dose. In comparison, in the IMPT plans, the range-limited six-axis correction was somewhat effective at improving the CTV coverage but did not contribute to reducing the CTV D2 or OAR dose. This may be related to the characteristics of the IMPT plans with robust planning and means that the effects of alignment distortion and changes in body shape that cannot be corrected by image registration are greater with IMPT plans than with VMAT plans, and that six-axis correction cannot correct them. Deterioration of the dose distribution due to anatomical changes was more severe in IMPT than in VMAT, consistent with the report of Góra et al. (17). Therefore, in head and neck IMPT, it is more important to correct vertebral alignment distortion by repositioning and to correct body thickness changes by ART.

The relationships between the position error of each anatomical structure and dose deterioration found in this study can be applied to optimization of the margin size. A previous study reported that anisotropic margins had lower margin size than isotropic margins (25). Therefore, it is possible that expanding the margin for the CTV around the vertebrae, clivus and mandible may help suppress deterioration of the CTV coverage. However, in the treatment of nasopharyngeal cancer, OARs such as the cochlea, brainstem and chiasm are in close proximity and margin expansion is often not possible, although there is a report that uncompromised IGRT is possible using an anisotropic margin generated with emphasis on the OARs (26). However, the CTV for head and neck cancer is generally broad and there are multiple OARs, which means that this method cannot be used. According to our results, the anisotropic margins at the initial planning may not be optimal because the positional variation of each anatomical feature increased with the number of treatment days and varied from patient to patient. Therefore, monitoring for vertebral alignment, and repositioning the patient and triggering the ART timing according to it are important to ensure the quality of dose distributions (27).

This study has several limitations. The number of cases examined was limited to 20 cases. The 10th, 20th, and 30th CT images included the effects of both positioning errors and changes in body shape, and it was not possible to separate these effects in the dose index evaluation. We used MIM Maestro® for automatic image registration, other automatic registration algorithms may provide different results.

Conclusion

We searched for anatomical structures that affect the dose indices in IMPT and VMAT. CTV coverage was intermediately correlated with positioning errors of vertebrae and OAR doses were strongly correlated with positioning errors of the C2, clivus, and mandible. In particular, the positional errors in the pitch and vertical directions of the vertebrae increased with the elapsed treatment days, and may be affected by changes in body shape. Six-axis correction improved the CTV coverage and OAR dose in VMAT.

Acknowledgements

The Authors wish to express sincere gratitude to Yoshihisa Muramatsu Ph.D. for supporting the research facilities and the environment for conducting this research. We thank Edanz (https://jp.edanz.com/ac) for editing a draft of this article.

Footnotes

  • Received January 17, 2023.
  • Revision received February 5, 2023.
  • Accepted February 9, 2023.

References

  1. ↵
    1. van de Water TA,
    2. Lomax AJ,
    3. Bijl HP,
    4. de Jong ME,
    5. Schilstra C,
    6. Hug EB and
    7. Langendijk JA
    : Potential benefits of scanned intensity-modulated proton therapy versus advanced photon therapy with regard to sparing of the salivary glands in oropharyngeal cancer. Int J Radiat Oncol Biol Phys 79(4): 1216-1224, 2011. PMID: 20732761. DOI: 10.1016/j.ijrobp.2010.05.012
    OpenUrlCrossRefPubMed
    1. Steneker M,
    2. Lomax A and
    3. Schneider U
    : Intensity modulated photon and proton therapy for the treatment of head and neck tumors. Radiother Oncol 80(2): 263-267, 2006. PMID: 16916557. DOI: 10.1016/j.radonc.2006.07.025
    OpenUrlCrossRefPubMed
  3. ↵
    1. Barten DL,
    2. Tol JP,
    3. Dahele M,
    4. Slotman BJ and
    5. Verbakel WF
    : Comparison of organ-at-risk sparing and plan robustness for spot-scanning proton therapy and volumetric modulated arc photon therapy in head-and-neck cancer. Med Phys 42(11): 6589-6598, 2015. PMID: 26520750. DOI: 10.1118/1.4933245
    OpenUrlCrossRefPubMed
  4. ↵
    1. Prabhakar R,
    2. Laviraj MA,
    3. Haresh KP,
    4. Julka PK and
    5. Rath GK
    : Impact of patient setup error in the treatment of head and neck cancer with intensity modulated radiation therapy. Phys Med 26(1): 26-33, 2010. PMID: 19576833. DOI: 10.1016/j.ejmp.2009.05.001
    OpenUrlCrossRefPubMed
    1. Siebers JV,
    2. Keall PJ,
    3. Wu Q,
    4. Williamson JF and
    5. Schmidt-Ullrich RK
    : Effect of patient setup errors on simultaneously integrated boost head and neck IMRT treatment plans. Int J Radiat Oncol Biol Phys 63(2): 422-433, 2005. PMID: 16168835. DOI: 10.1016/j.ijrobp.2005.02.029
    OpenUrlCrossRefPubMed
  6. ↵
    1. Liu W,
    2. Patel SH,
    3. Harrington DP,
    4. Hu Y,
    5. Ding X,
    6. Shen J,
    7. Halyard MY,
    8. Schild SE,
    9. Wong WW,
    10. Ezzell GE and
    11. Bues M
    : Exploratory study of the association of volumetric modulated arc therapy (VMAT) plan robustness with local failure in head and neck cancer. J Appl Clin Med Phys 18(4): 76-83, 2017. PMID: 28503916. DOI: 10.1002/acm2.12099
    OpenUrlCrossRefPubMed
  7. ↵
    1. Thomson DJ,
    2. Teo B-KK,
    3. Ong A,
    4. Ang KW,
    5. Kirk M,
    6. Ahn PH,
    7. Lukens JN,
    8. Swisher-McClure S,
    9. Liptrot T,
    10. Solberg TD,
    11. Slevin NJ and
    12. Lin A
    : The impact of anatomic change on pencil beam scanning in the treatment of oropharynx cancer. Int J Part Ther 2(2): 394-403, 2015. DOI: 10.14338/ijpt-15-00002.1
    OpenUrlCrossRef
    1. Cubillos-Mesías M,
    2. Baumann M,
    3. Troost EGC,
    4. Lohaus F,
    5. Löck S,
    6. Richter C and
    7. Stützer K
    : Impact of robust treatment planning on single- and multi-field optimized plans for proton beam therapy of unilateral head and neck target volumes. Radiat Oncol 12(1): 190, 2017. PMID: 29183377. DOI: 10.1186/s13014-017-0931-8
    OpenUrlCrossRefPubMed
    1. Paganetti H,
    2. Botas P,
    3. Sharp GC and
    4. Winey B
    : Adaptive proton therapy. Phys Med Biol 66(22): ac344f, 2021. PMID: 34710858. DOI: 10.1088/1361-6560/ac344f
    OpenUrlCrossRefPubMed
    1. Cubillos-Mesías M,
    2. Troost EGC,
    3. Lohaus F,
    4. Agolli L,
    5. Rehm M,
    6. Richter C and
    7. Stützer K
    : Quantification of plan robustness against different uncertainty sources for classical and anatomical robust optimized treatment plans in head and neck cancer proton therapy. Br J Radiol 93(1107): 20190573, 2020. PMID: 31778315. DOI: 10.1259/bjr.20190573
    OpenUrlCrossRefPubMed
  11. ↵
    1. Stützer K,
    2. Jakobi A,
    3. Bandurska-Luque A,
    4. Barczyk S,
    5. Arnsmeyer C,
    6. Löck S and
    7. Richter C
    : Potential proton and photon dose degradation in advanced head and neck cancer patients by intratherapy changes. J Appl Clin Med Phys 18(6): 104-113, 2017. PMID: 28921843. DOI: 10.1002/acm2.12189
    OpenUrlCrossRefPubMed
  12. ↵
    1. Pflugfelder D,
    2. Wilkens JJ and
    3. Oelfke U
    : Worst case optimization: a method to account for uncertainties in the optimization of intensity modulated proton therapy. Phys Med Biol 53(6): 1689-1700, 2008. PMID: 18367797. DOI: 10.1088/0031-9155/53/6/013
    OpenUrlCrossRefPubMed
    1. Unkelbach J,
    2. Bortfeld T,
    3. Martin BC and
    4. Soukup M
    : Reducing the sensitivity of IMPT treatment plans to setup errors and range uncertainties via probabilistic treatment planning. Med Phys 36(1): 149-163, 2009. PMID: 19235384. DOI: 10.1118/1.3021139
    OpenUrlCrossRefPubMed
    1. Fredriksson A,
    2. Forsgren A and
    3. Hårdemark B
    : Minimax optimization for handling range and setup uncertainties in proton therapy. Med Phys 38(3): 1672-1684, 2011. PMID: 21520880. DOI: 10.1118/1.3556559
    OpenUrlCrossRefPubMed
    1. van Dijk LV,
    2. Steenbakkers RJ,
    3. ten Haken B,
    4. van der Laan HP,
    5. van’t Veld AA,
    6. Langendijk JA and
    7. Korevaar EW
    : Robust intensity modulated proton therapy (IMPT) increases estimated clinical benefit in head and neck cancer patients. PLoS One 11(3): e0152477, 2016. PMID: 27030987. DOI: 10.1371/journal.pone.0152477
    OpenUrlCrossRefPubMed
  16. ↵
    1. Cubillos-Mesías M,
    2. Troost EGC,
    3. Lohaus F,
    4. Agolli L,
    5. Rehm M,
    6. Richter C and
    7. Stützer K
    : Including anatomical variations in robust optimization for head and neck proton therapy can reduce the need of adaptation. Radiother Oncol 131: 127-134, 2019. PMID: 30773179. DOI: 10.1016/j.radonc.2018.12.008
    OpenUrlCrossRefPubMed
  17. ↵
    1. Góra J,
    2. Kuess P,
    3. Stock M,
    4. Andrzejewski P,
    5. Knäusl B,
    6. Paskeviciute B,
    7. Altorjai G and
    8. Georg D
    : ART for head and neck patients: On the difference between VMAT and IMPT. Acta Oncol 54(8): 1166-1174, 2015. PMID: 25850583. DOI: 10.3109/0284186X.2015.1028590
    OpenUrlCrossRefPubMed
  18. ↵
    1. Lalonde A,
    2. Bobić M,
    3. Winey B,
    4. Verburg J,
    5. Sharp GC and
    6. Paganetti H
    : Anatomic changes in head and neck intensity-modulated proton therapy: Comparison between robust optimization and online adaptation. Radiother Oncol 159: 39-47, 2021. PMID: 33741469. DOI: 10.1016/j.radonc.2021.03.008
    OpenUrlCrossRefPubMed
  19. ↵
    1. Sonke JJ,
    2. Aznar M and
    3. Rasch C
    : Adaptive radiotherapy for anatomical changes. Semin Radiat Oncol 29(3): 245-257, 2019. PMID: 31027642. DOI: 10.1016/j.semradonc.2019.02.007
    OpenUrlCrossRefPubMed
  20. ↵
    1. Albertini F,
    2. Matter M,
    3. Nenoff L,
    4. Zhang Y and
    5. Lomax A
    : Online daily adaptive proton therapy. Br J Radiol 93(1107): 20190594, 2020. PMID: 31647313. DOI: 10.1259/bjr.20190594
    OpenUrlCrossRefPubMed
  21. ↵
    1. Nenoff L,
    2. Matter M,
    3. Hedlund Lindmar J,
    4. Weber DC,
    5. Lomax AJ and
    6. Albertini F
    : Daily adaptive proton therapy - the key to innovative planning approaches for paranasal cancer treatments. Acta Oncol 58(10): 1423-1428, 2019. PMID: 31364904. DOI: 10.1080/0284186X.2019.1641217
    OpenUrlCrossRefPubMed
  22. ↵
    1. Ding Z,
    2. Xiang X,
    3. Zeng Q,
    4. Ma J,
    5. Dai Z,
    6. Kang K and
    7. Bi S
    : Evaluation of plan robustness on the dosimetry of volumetric arc radiotherapy (VMAT) with set-up uncertainty in Nasopharyngeal carcinoma (NPC) radiotherapy. Radiat Oncol 17(1): 1, 2022. PMID: 34980178. DOI: 10.1186/s13014-021-01970-8
    OpenUrlCrossRefPubMed
  23. ↵
    1. Aoyama T,
    2. Shimizu H,
    3. Sasaki K,
    4. Ando M,
    5. Kaneda N,
    6. Tachibana H,
    7. Suzuki K and
    8. Kodaira T
    : A relationship between cervical vertebrae twisting and cranial angle in head and neck radiotherapy. In Vivo 34(5): 2401-2406, 2020. PMID: 32871765. DOI: 10.21873/invivo.12053
    OpenUrlAbstract/FREE Full Text
  24. ↵
    1. Otsuka M,
    2. Monzen H,
    3. Ishikawa K,
    4. Doi H,
    5. Matsumoto K,
    6. Tamura M and
    7. Nishimura Y
    : Variations of the dose distribution between CT- and CBCT-based plans for oropharyngeal cancer. In Vivo 33(4): 1271-1277, 2019. PMID: 31280218. DOI: 10.21873/invivo.11599
    OpenUrlAbstract/FREE Full Text
  25. ↵
    1. Guerreiro F,
    2. Seravalli E,
    3. Janssens GO,
    4. van de Ven CP,
    5. van den Heuvel-Eibrink MM and
    6. Raaymakers BW
    : Intra- and inter-fraction uncertainties during IGRT for Wilms’ tumor. Acta Oncol 57(7): 941-949, 2018. PMID: 29457751. DOI: 10.1080/0284186X.2018.1438655
    OpenUrlCrossRefPubMed
  26. ↵
    1. Radonic S,
    2. Besserer J,
    3. Bley CR,
    4. Schneider U and
    5. Meier V
    : A concept for anisotropic PTV margins including rotational setup uncertainties and its impact on the tumor control probability in canine brain tumors. Biomed Phys Eng Express 8(6), 2022. PMID: 35981496. DOI: 10.1088/2057-1976/ac8a9f
    OpenUrlCrossRefPubMed
  27. ↵
    1. Hirotaki K,
    2. Moriya S,
    3. Tachibana H and
    4. Sakae T
    : Detection of anatomical changes using two-dimensional x-ray images for head and neck adaptive radiotherapy. Med Phys 49(5): 3288-3297, 2022. PMID: 35235222. DOI: 10.1002/mp.15587
    OpenUrlCrossRefPubMed
    1. Brierley JD,
    2. Gospodarowicz MK and
    3. Wittekind C
    : TNM Classification of Malignant Tumours, Eighth Edition. Hoboken, New Jersey, John Wiley & Sons, Ltd, 2017.
Back to top

In this issue

Print
Download PDF
Article Alerts
Email Article
Citation Tools
Reprints and Permissions
Impact of Anatomical Position Errors on Dose Distribution in Head and Neck Radiotherapy and Robust Image Registration Against Anatomical Changes
KOUTA HIROTAKI, KENTO TOMIZAWA, SHUNSUKE MORIYA, MASASHI ITO, TAKEJI SAKAE
Anticancer Research Apr 2023, 43 (4) 1827-1834; DOI: 10.21873/anticanres.16336
Twitter logo Facebook logo Mendeley logo

Jump to section

Keywords