Three-dimensional conformal arc radiotherapy using a C-arm linear accelerator with a computed tomography on-rail system for prostate cancer: clinical outcomes

Along with radical prostatectomy and brachytherapy, external beam radiotherapy (EBRT) with or without hormonal therapy is one of the most commonly employed curative treatment options for localized prostate cancer [1‐3]. Recent technical developments have allowed conformal treatment to deliver both high-dose radiation to the target volume and reduced radiation to adjacent organs at risk, including the rectum and bladder [4, 5]. By “conformal treatment,” we mean both image-guided radiotherapy (IGRT) and high-technology EBRT, such as three-dimensional conformal radiotherapy (3D-CRT) and intensity-modulated radiotherapy (IMRT). As a consequence of these technical developments, greater effectiveness has been observed in large retrospective studies and several prospective randomized trials of treatment outcomes [6‐12]. In fact, dose escalation has improved local control and cause-specific survival [13‐15]. Moreover, clinical data from the Fox Chase Cancer Center (Philadelphia, PA, United States) have suggested that a dose–response relationship exists, even in the ultra-high dose range beyond 80 Gy [16]. Further, better disease control was obtained with ultra-high doses, as compared with standard doses [16].

Although relatively steep dose gradients can be applied in IMRT, dose escalation is believed to have limited applicability in 3D-CRT because of treatment-related morbidities. Several studies have shown correlations between higher 3D-CRT doses and increased risks of grade ≥2 late rectal and urinary toxicities [7, 12, 17].

At Sagamihara Kyodo Hospital (Kanagawa, Japan), image-guided three-dimensional conformal arc radiotherapy (3D-CART) has been implemented using a linear accelerator with a computed tomography (CT) on-rail system for localized prostate cancer since 2004. Dose escalation above 78 Gy (median dose: 82 Gy) has been investigated without the use of IMRT since 2006. The CT on-rail system was employed as a tool for IGRT. As compared with IMRT, our treatment demands less time, effort and cost for RT planning and verification because it does not require an inverse-planning approach. To the best of our knowledge, there have not been any published reports regarding the clinical results of 3D-CART for prostate cancer, as applied with similarly high doses and the use of a CT on-rail system as an IGRT tool in Japan. The aim of this retrospective study is to elucidate our method and report the treatment outcomes and toxicities observed in patients who received CT on-rail system-guided 3D-CART as a treatment for prostate cancer.

This study reports the clinical results of definitive treatment with CT-guided 3D-CART in patients with localized prostate cancer. The clinical outcomes and the treatment-related toxicity profile that we observed during the follow-up period (median, 61.6 months) appeared favorable in comparison with previous reports of IMRT [20‐22]. To the best of our knowledge, this is the only report regarding the use of CT on-rail system-guided 3D-CART to deliver high doses (median dose: 82 Gy) for patients with prostate cancer in Japan. Further, in previously published Japanese studies of EBRT for prostate cancer, the prescribed doses have all been under 80 Gy [23‐25].

With the development of EBRT, high-dose radiotherapy has become a well-established treatment for localized prostate cancer. In several randomized trials and a meta-analysis, comparisons of high- and low-dose RT have shown that high doses offer significant improvements in biochemical control [11, 12, 26]. Moreover, the findings of some studies have suggested that doses in the ultra-high range (beyond 80 Gy) offer clinical benefit in the form of disease control [16, 20, 27].

Despite the clinical benefits that are offered by dose escalation, the delivery of high-dose RT to the prostate is associated with elevated rates of GI and GU toxicities. The toxicities reported in previous studies of EBRT are summarized in Table 4. Studies that employed 3D-CRT with doses exceeding 74 Gy showed rates of treatment-related grade ≥2 GI and GU toxicities of 5.5–26 and 2.9–28 %, respectively [6‐8, 11, 12]. In a randomized trial of 3D-CRT from the M.D. Anderson Cancer Center (Houston, TX, United States), Kuban et al. showed that the rate of grade ≥2 late GI complications was twice as high in patients treated to 78 Gy than in patients treated to 70 Gy (26 % versus 13 %, respectively) [12]. Michalski et al. conducted a multi-institutional dose-escalation phase I–II study of 3D-CRT (RTOG 9406) and reported that grade ≥2 late GI and GU toxicities occurred in 25–26 and 23–28 % of patients, respectively, with a dose of 78 Gy in 39 fractions [8]. Reports of dose escalations above 74 Gy using IMRT have shown rates of grade ≥2 GI and GU toxicities of 2.4–18 and 3.5–22 %, respectively [20‐22]. In a retrospective analysis from the Memorial Sloan Kettering Cancer Center (New York, NY, United States), Spratt et al. reported that grade ≥2 late GI and GU toxicities occurred at incidences of 4.4 and 21.1 %, respectively, in patients who had undergone IMRT to a total dose of 86.4 Gy (1.8 Gy/fraction) [20]. Eada et al. showed that grade ≥2 GI and GU toxicities occurred in 2.4 and 3.5 % of patients, respectively, following IMRT with a total dose of 74–78 Gy (2 Gy/fraction) [21]. Comparing the complication rates associated with IMRT and 3D-CRT, GI toxicity rates are generally lower for IMRT [10, 28, 29].

In the current study, rates of grade 2 late GI and GU toxicities were 1.9 and 10.3 %, respectively, and rates of grade 3 late GI and GU toxicities were 0.4 and 1.1 %, respectively. There is a striking variation in the toxicity rates that have been reported across different studies. The large variation probably results from differences in treatment methods, radiation doses, and toxicity assessments. Notably, there is no single, standardized tool for measuring treatment-related toxicity. Even when we consider these between-study differences, our results for 3D-CART are comparable to those previously reported for IMRT, although we prescribed high doses (78–86 Gy) and employed a non-IMRT technique. There are several potential explanations for this result, as described below.

First, our approach reduced the volume of the rectum that was exposed to high-dose radiation, as compared with conventional 3D-CRT approaches such as the static five-to-seven port technique. In our 3D-CART approach, the rectal volume that was exposed to high doses (V80R, V70R) compares favorably with that of IMRT (Table 1 and Fig. 2). Our treatment planning method could have contributed to the low frequency of late rectal toxicities.

Second, we used a linac with CT on-rail system as an IGRT tool in daily treatment sessions. Consequently, we were able to confirm that the target was enclosed in the treatment field before every RT session.

Third, we were able to reduce the margins using this IGRT technique. IGRT is believed to be the best method of reducing margins. Sveistrup et al. [30] reported the toxicity rates in patients treated with image-guided IMRT and 3D-CRT without daily image guidance. Three hundred eighty-eight patients were treated at a dose of 78 Gy using IMRT, based on daily image guidance with fiducial markers. Further, 115 patients were treated at a dose of 76 Gy using 3D-CRT without daily image guidance. The 2-year rates of grade ≥2 GI toxicity were 5.8 and 57.3 % in the patients treated with image-guided IMRT and 3D-CRT, respectively (p < 0.001). Regarding GU toxicity, the corresponding values were 29.7 and 41.8 %, respectively (p = 0.011). In a study of IGRT and non-image guided RT, Zelefsky et al. [31] found that IGRT was associated with improved biochemical tumor control in high-risk patients, as well as a lower rate of late urinary toxicity in the total patient cohort.

In the present study, we set a 5-mm margin for the CTV in all directions, which is smaller than the margins reported previously [8, 20, 21, 23, 27]. As a tool for IGRT, the CT on-rail system provides better image quality than is obtainable using cone-beam CT, which is often equipped with the latest linacs. Therefore, we attempted to set up patients using prostate calcifications or the prostate itself as markers in daily IGRT, without implanting artificial fiducial markers (i.e., gold markers).

Fourth, we reduced the prescribed dose by 2–4 Gy in patients who were thought to have possible risk factors for late toxicities (i.e., use of anticoagulants, diabetes, or the occurrence of severe acute toxicities). This therapeutic strategy may also have reduced late toxicities. Considering the risk of toxicities due to the RT, however, 78 Gy remains a high dose for patients with such co-morbidity factors. In hindsight, lower doses should have been used in such patients.

A small PTV margin could leave the target out of the radiation fields, worsening the results of treatment. In our study, however, the 5-year FFBF rates for low-, intermediate-, and high-risk patients were 100, 91.8, and 90.3 %, respectively. Although 15 patients died during the follow-up period, no patient died of prostate cancer and the 5-year OS rate was 93.5 %. These outcomes compare favorably with those of previous reports on IMRT. Our results indicate that high-accuracy positioning with IGRT can allow small PTV margins to be used in EBRT for prostate cancer.

As compared with IMRT, our therapy demands less time and effort for RT planning and verification because it does not require an inverse-planning approach. In order to ensure accurate dose delivery, IMRT requires meticulous implementation and execution with rigorous quality assurance and control. In Japan, two or more full-time radiation oncologists and one or more full-time radiation physicists are needed for IMRT at each radiation therapy facility. However, many institutions do not fulfill these criteria owing to a lack of human resources, and they are therefore not allowed to implement IMRT. At our institution, we performed the 3D-CART technique routinely in daily RT sessions for patients with localized prostate cancer patients, without encountering any difficulty. This method saved time during preparation before RT—in fact, RT began two days after planning CT imaging. Considering the clinical outcomes and ease of preparation before RT, the reported approach may provide another treatment choice for prostate cancer patients.

The key limitations to this study are its retrospective nature and the differences in the toxicity assessment methods that were employed by each radiation oncologist. However, it should also be considered that few patients suffered from severe late treatment-related toxicity during the follow-up period (median duration, 61 months), despite the application of high-dose RT (78–86 Gy). Further, the initial treatment results are comparable to previous reports of IMRT dose escalation studies, although a direct comparison between the 3D-CART approach and IMRT would of course be necessary to reach definitive conclusions. Longer durations of follow-up will be needed to provide a full evaluation of the treatment results, especially for patients who received adjuvant hormonal therapy for long periods. Additionally, given the inclusion of low-risk patients and the absence of deaths from prostate cancer, the study’s results could have been affected by overdiagnosis to some extent, since the outcomes of overdiagnosed patients would not reflect the actual benefits of RT.

In conclusion, this report confirmed the feasibility and effectiveness of high-dose (≥78 Gy) conformal arc radiotherapy in a large cohort of patients with prostate cancer. The treatment outcomes were comparable to those that have previously been reported for IMRT. In this retrospective study, we found that the 3D-CART approach with IGRT using a CT on-rail system was associated with a low frequency of late rectal toxicities.