Dosimetric Comparison of Static Field Intensity-Modulated Radiotherapy and Volumetric Modulated Arc Therapy for Adjuvant Treatment of Patients with Endometrial Cancer

VMAT techniques were developed after implementation of IMRT. Over the past ten years they were compared to IMRT and 3DCRT, Wong et al., compared VMAT, IMRT and 3DCRT.[,] The study included 5 post-operative endometrial cancer cases, some of them treated with para-aortic volumes as well. The arc technique consisted of 30-300° and 60-330° gantry arrangements and the IMRT was delivered with 8 fields. There were no significant difference between IMRT and VMAT. Both of them were superior to 3DCRT regarding the small bowel and iliac BM doses. Unlike that, we used 7-field IMRT setup and did not restrict the angular arrangement of VMAT. With the use of static IMRT we observed significant reduction in doses to small bowel (Dmax: 48.35 vs. 49.09 Gy, D2%: 46.27 vs. 46.94 Gy, D2cc: 47.5 vs. 48.22 Gy, V45: 180.97 vs. 219.31 cc) and iliac BM (D2%: 46.93 vs. 47.45 Gy, V10 Gy: 95% vs. 99%) compared to VMAT. In our study, formula defined by RTOG was used for HI calculation (maximum isodose volume/reference isodose volume) and HI was better with static IMRT over VMAT (1.06 and 1.07, respectively). Wong et al., used a different formula (dose difference between D5% and D95% of PTV) and their results were in favor of IMRT against VMAT as well (7.5% dose difference vs. 11%, respectively).[]

Cozzi et al., compared single-arc VMAT and IMRT in 8 cases with cervix cancer who were treated with chemo-radiation.[] The prescription was 50.4 Gy in 28 fractions to PTV. Results were in favor of static IMRT both for CI and the doses to OAR. They reported significant reduction in V40 Gy of small bowel and "small bowel-PTV" with VMAT in contrast to our study which resulted with better Dmax, D2%, D2cc of small bowel and D2cc, V45 Gy of "small bowel-PTV" with IMRT (Table 3). We could not find any difference in parameters of bladder whereas Cozzi et al., reported reduction in D2% of "bladder-PTV", V40 Gy of bladder and "bladder- PTV" with VMAT.[12] Both studies showed reduction in V40 Gy of rectum, V40 Gy and D2cc of "rectum-PTV" with VMAT (51% vs. 59%, 12% vs. 21% and 40.66 Gy vs. 42.22 Gy respectively in our study). Although there is a decrease in dose to femoral heads in our study, it does not have any clinical importance.

For the parameters of PTV, Cozzi et al., reported no difference in D98% and V95% between two techniques but a superiority of VMAT for in D2%.[] In our study parameters representing maximal doses like D2%, D5%, V107% and V105% were lower with IMRT. On the other hand, with higher D98%, PTV V45 Gy, CTV V45 Gy and CTVmin, VMAT was capable of delivering better dose coverage (Table 2). We think that the main underlying reason for this distinction is the difference of PTV and anatomy between cervical and post-operative endometrial cancers. Additionally, it should be noted that Cozzi et al., did not define the pelvic BM as OAR and used another formula than RTOG for the calculation of CI and HI.

Yang et al., also compared 3DCRT, IMRT and a conformal double-arc technique on 10 cases with postoperative endometrial cancer.[] They made plans with an optimization for 50 Gy to ≥95% of the PTV. Regarding OAR parameters, arc plans had better results than 3DCRT and worse results than IMRT plans. But the arc technique used in this study was a rotational conformal modality with neither inverse planning nor intensity modulation. Mean MU for 3DCRT, arc and IMRT was reported as 240, 451 and 877 MU, respectively. Our results for MU/fraction and "beam on time" (IMRT/VMAT: 1688.58/465.5 MU and 337.72/166.42 seconds) are in concordance with the results of Cozzi et al., showing a marked advantage of VMAT.[] Risk of secondary malignancy due to higher MU of IMRT versus slightly higher toxicity risk with VMAT (except for rectum) is an issue open to discussion.

A recent dosimetric study by Sharfo et al., compared different IMRT and VMAT strategies on 10 cervix cancer patients using automated in-house planning software.[] The planning goal was to achieve highly conformal plans while sparing the OARs with a higher priority for the small bowel. The treatment delivery time was shorter with VMAT, but with IMRT superior plan quality was observed. Although performed on cervix and not on endometrial cancer patients, for us the results of this original work have profound importance, since they depend on an unbiased fully automated software solution where modern techniques were compared. In their conclusions, the authors also emphasized the importance of the trade-off between the plan quality and treatment delivery.

Two articles were published comparing 3DCRT, static field IMRT and helical tomotherapy using planning CT data of 10 endometrial cancer patients.[,] An interval from 95% to 110% was used for target coverage optimization. Lian et al., reported the results of extended-field radiotherapy for stage IIIC patients.[] Target coverage was identical in both IMRT and tomotherapy plans. No difference in integral dose between tomotherapy and IMRT was shown. Yang et al., reported statistically insignificant differences in conformity among IMRT and tomotherapy plans.[] Tomotherapy had 3.3% higher integral dose to normal tissue than static field IMRT (p<0.01).

In the present study, we could reach most of our dosimetric targets. Regarding the parameters of OAR, the objectives for bladder and femoral heads were achieved. Both techniques failed to keep V10 Gy of BM under 90%, but the V20 Gy target was easily reached. However it is important to emphasize for this group of patients that no hematologic toxicity is observed in our clinical routine. It was an unrealistic objective trying to keep V40 Gy of rectum under 35% because of the large intersection of PTV and rectum. For the same reason, it prohibited achieving dosimetric goals for small bowel except for the maximum dose constrain. We think that the main reason behind this problem is our choice of +1 cm margin around CTV to create PTV. This relatively large margin could be reduced with daily instead of weekly Cone Beam CT (CBCT). The choice between a smaller PTV margin with high body kV X-ray exposure due to daily CBCT versus a larger margin with lower exposure due to less frequent CBCT is beyond the scope of this study.[]

Even there is statistically significant difference between two techniques, one may find it as clinically unimportant. Thus, we did not observe any difference in toxicity profile during or after the treatment with IMRT or VMAT for post-operative endometrial cancer patients in our clinic so far. Acute toxicity is limited with genitourinary and gastrointestinal toxicity never higher than RTOG grade 2. Because of shorter treatment times compared to static IMRT, VMAT allows more time for patient set-up, daily image guidance and provides better patient comfort. Shorter treatment times also restrict intrafractional organ movement.[]

As limitations, our study contains 12 pairs of plans for matched comparison, and although unique for endometrial cancer, the findings we present are not surprising based on previously published data on cancers of uterine cervix and anal canal. We also want to emphasize that statistical correction of p value due to multiple hypothesis testing was performed neither in any of the previously published studies nor in our study. If we were to interpret our results with a strongly conservative Bonferroni correction, results with p>0.0008 should be discarded. This would leave only the differences in MU/fraction, beam on time and BM-PTV D2% parameters as statistically significant.