A dosimetric comparison of the use of equally spaced beam (ESB), beam angle optimization (BAO), and volumetric modulated arc therapy (VMAT) in head and neck cancers treated by intensity modulated radiotherapy

Abstract Introduction Previous studies have shown that the beam arrangement had significant influence on plan quality in intensity modulated radiotherapy (IMRT). This study aimed to evaluate the dosimetric performance of beam arrangement methods by employing equally spaced beams (ESB), beam angle optimization (BAO), and volumetric modulated arc therapy (VMAT) in the planning of five types of head and neck (H&N) cancers treated by IMRT. Methods Five plans of different beam arrangement methods were optimized for 119 H&N cancer patients with the prescription of 66–70 Gy for high‐risk planning target volume (PTV), 60 Gy for intermediate risk PTV, 54 Gy for low‐risk PTV using a simultaneously integrated boost method. The five‐beam arrangement methods were: ESB, coplanar BAO (BAOc), noncoplanar BAO (BAOnc), two‐arc VMAT (VMAT2), and three‐arc VMAT (VMAT3). The H&N cancers included cancers of nasopharynx, oral cavity, larynx, maxillary sinus, and parotid. Although the partial arc VMAT could be used in cases where the PTVs were situated at one side of the head such as the parotid, this arrangement was not included because it was intended to include only the beam arrangements that were applicable to all the types of head and neck cancers in the study. The plans were evaluated using a “figure‐of‐merit” known as uncomplicated target conformity index (UTCI). In addition, PTV conformation number and homogeneity index, normal tissue integral dose, and organ at risk (OAR) doses were also used. The mean values of these parameters were compared among the five plans. Results All treatment plans met the preset dose requirements for the target volumes and OARs. For nasopharyngeal cancer, VMAT3 and BAOnc demonstrated significantly higher UTCI. For cancer of oral cavity, most beam arrangement showed similar UTCI except ESB, which was relatively lower. For cancer of larynx, there was no significant difference in UTCI among the five‐beam arrangement methods. For cancers of maxillary sinus and parotid gland, the two BAO methods showed marginally higher UTCI among all the five methods. Conclusion Individual methods showed dosimetric advantages on certain aspects, and the UTCI of the BAO treatment plans are marginally greater in the case of maxillary sinus and parotid gland. However, if treatment time was included into consideration, VMAT plans would be recommended for cancers of the nasopharynx, oral cavity, and larynx.

Conclusion: Individual methods showed dosimetric advantages on certain aspects, and the UTCI of the BAO treatment plans are marginally greater in the case of maxillary sinus and parotid gland. However, if treatment time was included into consideration, VMAT plans would be recommended for cancers of the nasopharynx, oral cavity, and larynx. Intensity modulated radiotherapy (IMRT) has been the main treatment modality for many head and neck cancers due its relatively high target dose conformity and steep dose gradient at targetnormal tissue interfaces compared with the conventional three-dimensional conformal radiotherapy (3DCRT). While the dose distribution in IMRT is largely controlled by the beam modulation using dynamic multileaf collimators (MLC), the beam arrangement including beam number and beam angle employed in the treatment plan have been reported to have significant dosimetric influence in the plan quality in IMRT of many cancers including oesophagus, 4 lung, 5 pharynx and larynx, 6 and nasopharynx. 7 Equally spaced beam (ESB) arrangement has been commonly used in the early application of IMRT in head and neck cancers after replacing 3DCRT in the early nineties. [8][9][10] Volumetric modulated arc therapy (VMAT) and beam angle optimization (BAO) in the Eclipse treatment planning system (Varian Medical System, Palo Alto, USA) are the two more recent options in assigning IMRT beams. VMAT is the delivery of IMRT using rotating arc beams, 11,12 while BAO is the use of a specific optimization algorithm to select the optimum angles of static beams, either coplanar or noncoplanar.
Previous dosimetric studies on the applications of VMAT and ESB in head and neck cancers 13,14 reported that dual arc VMAT improved the target coverage and OAR sparing in cancers of oropharynx, hypopharynx, and larynx 6,15 and VMAT produced similar plan quality as ESB arrangement with marked reduction of monitor unit (MU) and shorter treatment delivery time. 13,14 Studies on BAO are limited. Some of them reported that coplanar BAO arrangement when applied to glioblastoma, prostate, and pancreatic cancers resulted in similar plan quality as ESB arrangement with reduced MU and number of fields. 16,17 Aside from coplanar beams in IMRT, BAO can also generate noncoplanar beam arrangement. Although noncoplanar IMRT has been reported to reduce the doses to OARs and normal tissues in prostate cancer patients, 18 its use in the clinical department is uncommon mainly due to the treatment setup inconvenience. However, with the recent emergence of 4Pi radiotherapy with compatible linear accelerators, 19 it is expected that the use of noncoplanar IMRT will be increased and its potential advantages can be better exercised.
To date, studies on IMRT beam arrangement for head and neck cancers have been limited to specific sites or just any two of the beam arrangement techniques. The optimum beam arrangement, in terms of dosimetric quality, for individual sites of head and neck cancers remains uncertain. Therefore we aimed to conduct a more comprehensive study that evaluated the dosimetric performance of five main IMRT beam arrangement methods on five types of common head and neck cancers that covered the various sites of this body region.

| MATERIALS AND METHODS
A total of 119 adult head and neck cancer patients treated by radical IMRT were randomly selected. They included cancers of the nasopharynx, oral cavity, larynx, parotid gland, and maxillary sinus.
Each cancer type consisted of a sample size of 25 except for maxillary sinus, which had 19 due to the limited number of cases available in the clinical department. The distributions of the T and N stages of the patients in each cancer group are summarized in Table 1.
where V T, ref = volume of target receiving a dose equal to or greater than the reference dose, V T = volume of target, V ref = volume receiving a dose equal to or greater than the reference dose.
The OARs considered for dosimetric comparison were the spinal cord, brain stem, and parotid glands (contralateral only for the parotid cancer group) as they were the relatively more critical organs and small changes in dose level would affect the risk of complications.
For the spinal cord and brain stem, D 2% was used for dose recording, whereas for the parotid gland, D 50% and D mean were used. For other OARs such as the optic nerve, cochlea, and pituitary gland, it was expected that they received relatively lower doses; slight differences would not have clinical significance and therefore were not included in this study. The normal tissue dose, which was expressed as the integral dose (in Gy*cm 3 ), was calculated by multiplying the D mean of the patient body included while planning CT scan excluding PTVs with the volume of this body region (normal tissue). 23 In addition, a "Figure of merit", also known as the uncomplicated target conformity index (UTCI), was used to rank the overall plan quality. 24 It was calculated by CN × Penalty of organs at risk (P OAR ) × Penalty of integral dose (P ID ). The higher the score, the better was the overall plan quality. The CN component of the UTCI was adopted from the Van't Riet study. 22 The P OAR and P ID were calculated by e À0:05 DiÀDtol ð Þ ½ with D i representing the actual received dose and D tol representing the tolerance dose. Since there was no established tolerance for integral dose, the D tol in the calculation of P ID was taken as the lowest achieved integral dose within the group of the same cancer. The role of P OAR and P ID in the equation was to penalize the UTCI score when the actual OARs of normal tissue dose exceeded their tolerance dose.
Statistical analysis was performed using the SPSS version 20 (IBM Corp, Armonk, NY). All the dose parameters and the UTCI scores were first tested for normality using the Shapiro-Wilk test.
The mean values of the dose parameters and UTCI scores for each beam arrangement group were calculated and compared. One-way repeated measures ANOVA test was used to analyze the differences among the five-beam arrangement methods. When there was significant difference among them, post hoc Tukey test was applied to further determine the ranking of each method.

3.A | Cancer of nasopharynx
For PTVH, ESB demonstrated the lowest HI and highest CN (  parameters showed relatively small differences and did not reach statistical significance. For the OARs, there was no significant difference in the spinal cord and brain stem doses. For the parotid gland and normal tissues, the two BAO methods (BAOc and BAOnc) gave relatively lower doses and the difference between them was minimal. With regard to UTCI, BAOc, BAOnc, VMAT2, and VMAT3 achieved similar scores, which were higher than that of the ESB.

3.C | Cancer of larynx
There was no significant difference in most of the dose parameters for the target volumes except that VMAT2 and VMAT3 gave a relatively higher CN in the PTVH (Table 4). Similar results were demonstrated in the OAR doses in which no significant difference was observed. For the normal tissues, the two BAO methods showed the lowest dose. However, there was no significant difference in UTCI among the five-beam arrangement methods.

3.D | Cancer of maxillary sinus
In general, little difference was observed in PTVs except that the two BAO methods showed relatively higher CN in PTVH and PTVI, and VMAT3 demonstrated the lowest HI for PTVL (Table 5). There was no significant dose difference for the OARs among the fivebeam arrangement methods except for the mean parotid dose, in which the two VMAT plans were relatively higher. For the normal tissues, the two BAO methods demonstrated relatively lower doses.
Furthermore, the UTCIs of the two BAO methods were significantly higher than the other three methods.

3.E | Cancer of parotid gland
There was no significant difference among the five-beam arrangement methods for dose parameters of the PTVs and OARs ( Table 6).
The two BAO methods delivered the lowest doses to the normal tissues and achieved the highest UTCI.

| DISCUSSION
Each of the beam arrangement methods had their uniqueness in delivering the tumoricidal dose to the tumor. The equally spaced beam in the ESB method directed the IM beams evenly from all angles around the patient and was considered the best in the T A B L E 2 Comparison of PTVs, OARs dose parameters, and integral dose between IMRT plans of five-beam arrangements for cancer of nasopharynx (n = 25). Because of this, they were expected to deliver higher integral dose to normal tissues. 25 The VMAT3 had the potential to produce more conformal dose distribution than VMAT2 but required one additional gantry rotation and therefore increased the treatment time. In this study, the BAO methods used 5-9 beams directed from selected angles. Beams that did not have contribution to the plan were elimi-

4.A | Cancer of nasopharynx
With regard to the target conformity and homogeneity, ESB performed better in PTVH and PTVI. This could be due to the fact that these target volumes were relatively less irregular in shape than the PTVL and, being more centrally situated at the skull, the evenly distributed intensity modulated IM beams were able to produce relatively more conformal dose distribution. However, for PTVL which extended to both sides of the neck, and was more irregular with an inverted U-shape (Fig. 1), the BAOnc and VMAT plans demonstrated relatively better dose coverage. The main reason was that with the use of noncoplanar beams in BAO and the greater number of effective beam angles from VMAT, they were both more effective in creating conformal high-dose volumes covering the irregular target. By the same argument, the OARs were better spared by these two beam arrangement methods. 18 By combining the performance in the target volumes and OARs, it was logical to see that both BAOnc and VMAT3 achieved relatively better plans among the five-beam arrangement methods. Overall, VMAT3 would be recommended because it would have a much shorter treatment delivery time compared to BAOnc.

4.B | Cancer of oral cavity
In terms of dose coverage to the target volume, VMAT plans in general performed better among the five-beam arrangement methods as it demonstrated the highest CN in all the PTVs. Since the oral cavity was a relatively large structure, tumors could arise from different locations in the oral cavity ranging from the periphery to the center.
The results showed evidence that the VMAT beam arrangement was more flexible to deal with target volume location variation, and its average performance on target coverage was better than the other T A B L E 5 Comparison of PTVs, OARs dose parameters, and integral dose between IMRT plans of five-beam arrangements for cancer of maxilla sinus (n = 19

4.E | Cancer of parotid gland
Target volumes of the parotid gland tumor were usually followed a triangular shape and would not poses great difficulty to the various beam arrangements. This was reflected in the dosimetric results of the target volumes in which there was no significant difference in all the dosimetric parameters among the five-beam arrangement methods. Similar to the maxillary sinus cancer, parotid tumors are situated at the lateral aspect of the head, and this would be a disadvantage for the ESB and VMAT beam arrangement. Relatively higher doses were found in the contralateral parotid gland in these plans although the differences did not reach statistical significance. Similar to the explanation for the maxillary sinus cancer, the integral dose in the BAO plans were significantly lower, and this led to overall better UTCI scores in these two plans. This echoed the study by Yirmibesoglu et al 28 who reported that four-field ipsilateral IMRT techniques provided excellent coverage while maximally sparing the contralateral parotid gland and submandibular gland. As a result, BAOc plans would be recommended as it achieved the same plan quality as the BAOnc but offered simpler treatment setup procedure.
Furthermore, by the same argument as stated for the cancer of maxillary sinus, the results of the present study were not against the use of partial arc VMAT in the cancer of parotid gland.

| CONCLUSION
The five-beam arrangement methods produced acceptable plans for all the five groups of head and neck cancer patients. Partial arc VMAT was not included in the beam arrangement methods because it was not commonly applied in centrally located cancers, therefore it could not be used for comparison among the five groups of head and neck cancers. The results showed that individual methods produced dosimetric advantages on certain aspects, and the UTCL scores were marginally greater in the BAO method in the cancers of the maxillary sinus and the parotid gland. However, if the treatment time was included into consideration, VMAT plans would be recommended for cancers of nasopharynx (VMAT3), oral cavity, and larynx (VMAT2).

CONF LICT OF I NTEREST
The authors declare no conflicts of interest with respect to the content of this manuscript.