Consistency Study of Multiparametric Magnetic Resonance and Computed Tomography Targeting in Targeted Radiotherapy for Prostate Cancer

Background: To compare the difference of location by computed tomography (CT) and multiparametric magnetic resonance imaging (mpMRI) on the target delineation and volume for organs at risk (OARs) among patients with prostate cancer. Methods: T1-weighted imaging (T1WI), T2-weighted imaging (T2WI), and CT were performed among 11 patients who received radiotherapy for prostate cancer at our center between August 2018 and December 2019. The target areas were delineated using the Eclipse system, and the radiotherapy plans were made based on the treatment planning system (TPS) to compare target volume and dose-volume histogram (DVH) relative to rectum and bladder. Results: The clinical target volume (CTV) of T1WI and T2WI decreased by 18.8% (P=0.001) and 22.72% (P=0.003), respectively, compared with CT. The planning target volume (PTV) on T1WI and T2WI were 20.45% (P=0.015) and 22.31% (P= 0.008) smaller than that dened by CT. There was no signicant difference in either CTV or PTV between the areas outlined on T1WI and T2WI. The DVH resulting from CT and MRI comparisons showed that the rectum and bladder dose levels were lower with MRI images compared with CT. It should be noted that at the lateral directions, the range of outlining on T2WI sequence were signicantly smaller than others. Conclusion: Target based on levels

increase the radiation dose to the target area. (3,6) Clinicopathological examination indicated that MRI could achieve a CTV accuracy of close to 95%. (7) T1WI can illustrate the external outline of the prostate gland, distinguishing the prostate from the pelvic muscles and periprostatic adipose tissue; however, imaging of the lesions within the gland is challenging, hindering the diagnosis of lymph node and bone metastases. (8-10) T2WI provide excellent soft-tissue contrast, allowing for the assessment of peritoneal involvement, and invasion in seminal vesicles and other adjacent tissues; (9)(10)(11) hence, T2WI MRI is currently the most promising method for PCa imaging. (2) High intra-target dose delivery in multifocal, non-uniform prostate tumors can improve local tumor control and patient prognosis. (2,11) According to the national comprehensive cancer network (NCCN) guidelines, delivery of 70-70.2 Gy in doses of 2.5-2.7 Gy is recommended for patients with PCa. (12) Since image quality affects the accuracy of target volume planning, improving image quality is critical for the success of radiotherapy. MRI has several advantages compared to CT, including low radiation, higher soft-tissue resolution, (13,14) and better discrimination of prostate lesions from the adjacent non-malignant tissues; (15) thus, MRI has emerged as a promising imaging modality for EBRT. (3,4,6) In this study, we assessed the performance of MRI in target volume identi cation and its potential to improve PCa radiotherapy outcomes.

Patient Information
A total of 11 patients with pathologically con rmed prostate cancer were selected at our center from August 2018 to December 2019. All patients were diagnosed with middle-risk to high-risk prostate cancer (group). The average age of these patients was 67 years, their T stages ranged from T2 to T4, their Gleason score ranged from 8 to 10, and their prostate-speci c antigen (PSA) levels were 0.039-248.6 ng/ml (Table 1). All patients received radiotherapy at Shandong Cancer Hospital following complete pelvic enhanced CT examination.

Target Area Delineation And Dose Calculation
Following the Radiation Therapy Oncology Group (RTOG) prostate cancer target area delineation guidelines, one experienced radiotherapy experts delineated the target areas on CT and MRI, respectively.
For each patient, a physicist used the Eclipse 8.6 (Varian Medical Systems Inc., Palo Alto) planning system to make a radiotherapy plan. All patients received 7-eld IMRT. The CTV included the entire prostate (mass and normal tissue) and the seminal vesicles within 2 of the top of the prostate. All directions on CTV are expanded by 5 mm to become PTV. Organ at risk (OARs) includes bladder, rectum and bilateral femoral heads. The target area dose calculation of CT positioning and MRI positioning was performed on CT and MRI-CT fusion images, respectively, and the calculation was based on the location of bone markers. Each patient was irradiated with 6MV X-ray. The DT (Total dose) was 70-76Gy, 1.8-2Gy/fraction.
For the three groups of radiotherapy plans, the dose parameters of the OARs (bladder, rectum) were extracted from DVH. The geometric center of the prostate cancer target area was de ned as point O, and the horizontal plane of that point was included in the analysis. OA was de ned as the distance from point O to the front end of the target area, and OP was the distance from point O to the back end. OR and OL were the distances to the right and left ends of the target area, respectively, in the same horizontal plane as the geometric center of the target area. The apex and base of the prostate were also labeled.

Statistical Analysis
All analyses were performed using SPSS Statistics 22.0 statistical software, and the measurement data were expressed as X̅ ±SD. All data were subjected to the Shapiro-Wilk normality test, followed by the t-test  (Fig. 1, Table 2), respectively. Compared with CT imaging, the CTV volumes delineated by T1WI and T2WI were reduced by 18.9% (P = 0.001), 23.63% (P = 0.003), respectively. The difference was statistically signi cant, that is, the CTV volumes delineated by MRI were signi cantly smaller than those delineated by CT. However, there was no signi cant difference between T1WI and T2WI (P = 0.323) ( Table 1, Fig. 2).  (Table 2), respectively. Compared with CT imaging, the PTV delineated by T1WI and T2WI were reduced by 20.45% (P = 0.005), 22.86% (P = 0.008), respectively. The difference was statistically signi cant, that is, the PTV volumes delineated by MRI was signi cant less than those delineated by CT ( Table 2, Fig. 2). PTV volumes delineated by T1WI and T2WI showed no signi cant difference (P = 0.527) ( Table 2, Fig. 2). On a box-and-whisker plot, the PTV box was signi cantly higher for CT than for T1WI or T2WI, and the upper section of the box was relatively short, indicating that the PTVs were greater on CT than on the MRI sequences and that there were more patients with high values on CT. Therefore, the PTVs delineated by T1WI and T2WI were signi cantly more speci c than those delineated by CT (Fig. 2).

Dose Comparison Of Oars
When we compared the 7-eld IMRT planning of the T1WI, T2WI and CT groups, the bladder exposure dose volumes of V20, V25, V30, V35, V40, V45, V50, V55, V60, V65 and V70 in the rst two groups were smaller than those of the CT group (Fig. 3), but the differences among the three groups were not signi cant (P > 0.05) ( Table 3, Table 4). As for the rectal dose, the values of exposure dose/total volume × 100% were not signi cantly different among the three groups when the exposure dose was < 30 Gy, while that of the CT group was slightly smaller than those of the T1WI group or the T2WI group when the exposure volume was > 30 Gy, but the difference between the two groups was not statistically signi cant (P > 0.05) ( Table 3, Table 4, Fig. 3).  Data analysis showed that except for the OR and OL directions in the T2WI group, which had signi cantly smaller values than those of the CT group (P = 0.002, P = 0.037) (Table 5), the values of the remaining positions showed no signi cant difference between the CT and T1WI groups, between the CT and T2WI groups, or between the T1WI and T2WI groups (Fig. 4, Table 5). The average differences in apex between T2WI and CT and between T1WI and CT were 0.72 ± 1.37 cm and 0.33 ± 0.22 cm, respectively, but these differences, despite their magnitude, were not signi cant (P = 0.220, P = 0.840). The average differences in base of prostate between T2WI and CT and between T1WI and CT were0.64 ± 0.89 cm and 0.25 ± 0.22 cm, respectively; once again, these differences, although pronounced, were not signi cant (P = 0.848, P = 0.146). Overall, the PTV location of CT, T1WI, and T2WI in all positions were approximately the same.
Although the PTV of CT was larger than MRI in all directions, there was no statistically signi cant difference.

Disscussion:
In this study, we compared the target volume and localization as well as the exposure dose of OAR with CT and mpMRI. The results showed that the CTV delineated by T1WI and T2WI were 81.1% and 76.37%, smaller than the same target de ned by CT, respectively. There was no signi cant difference in target volume between T1WI and T2WI. The PTV volumes delineated by T1WI and T2WI were 79.55% and 76.14% of the volume measured through CT, respectively. Target localization showed an obvious disparity in the apex and base of the prostate. The average differences in apex contouring of PTV between T1WI and T2WI were 0.64 ± 0.89 cm and 0.25 ± 0.22 cm, compared with CT, respectively, and the average distances of these two at the base were 5.64 ± 0.89 cm and 5.25 ± 0.22 cm, respectively. Concurrently, we found that the PTVs delineated by CT and T2WI in the lateral (OR and OL) directions were signi cantly different (P = 0.002 and 0.037, respectively); the average differences were 0.49 ± 0.63 cm and 0.52 ± 0.89 cm, respectively. By comparing the DVHs of the IMRT planning, we found that the two mpMRI groups received slightly lower bladder and rectal doses than the CT group, but the difference was not statistically signi cant.
Radiotherapy is an important treatment for prostate cancer and is widely used in decision-making for prostate cancer treatment at various stages. (11,16) Currently, EBRT is the main radiotherapy for prostate cancer .(4) CT is commonly used in image acquisition and delineation to support clinical decision making. Although CT has the advantages of rapid operation, accurate positioning, clear imaging. This image technology can provide high tissue density resolution in the formulation and implementation, especially electronic density information provided for radiotherapy dose calculation. However, it has a notable weakness in that it does not su ciently display the internal structure and edge of the prostate, especially at the apex of the prostate and the junction of the prostate, rectum and bladder. (14,16,17) The soft tissue resolution of CT is poor, and it is di cult to distinguish apex from the surrounding soft tissues considering this area is wrapped by distal urethra, upper penis bulb and the levator ani muscle. This area appears only as a homogenetic soft tissue density in the CT image. This imaging defect This often makes the delineation of the radiotherapy target area inaccurate. The European society for radiotherapy and oncology(ESTRO) guidelines indicate that the delineation range of the apex of the prostate should be an area 1 above the penis bulb.
(2) Moreover, series of studies have con rmed that the target area determined by CT is much larger than the actual tumor distribution range, which makes it di cult to avoid toxic and side effects in radiotherapy, thereby reducing the quality of life of patients. (6) The high resolution of MRI soft tissue can clearly display the internal structure of the prostate and the boundaries of pelvic organs such as the rectum and bladder. Even the complex structure of the apex of the prostate can be clearly distinguished, especially in the T2WI sequence, which signi cantly reduces the observer The inter-error is gradually applied to the target area of prostate cancer. (9,16,17) Various literatures have con rmed that the prostate volume delineated on CT is larger than that on MRI. Interestingly, a signi cantly differentness in the lateral (OR and OL) position is noticed,which is consistent with the report of Sannazzarl et al. (19). This result may be caused by the superior soft tissue contrast of MRI in prostate anatomical imaging (see Fig. 5) .Because anterior of the prostate is xed by the anterior bromuscular stroma and the posterior prostate has the thickness capsule ,the border in the anteriorposterior direction is visible easily on both images (22). Considering the distinct anatomical characteristics, the PTV delineation showed no signi cant difference in anterior-posterior direction. In the lateral direction of prostate, however, shown a different circumstance. The levator ani muscle, which xed prostate, is thickest in lateral direction, resulting in the signi cant discrepancies between CT and MRI, particularly in T2WI sequence.
In this study, the target volume reduced by MRI was smaller than previously reported, which may be due to the application of more accurate MRI equipment, with a layer spacing of 3 mm instead of the previous 8 mm, and the improvement of imaging compared with previous studies. In this study, compared with CT, the delineation range of prostate apex by MRI positioning was 0.33-0.72 cm, and that of the prostate bottom was 0.25-0.64 cm. This suggests that clinicians can appropriately reduce the target area range of upper and lower levels of the lesion when delineating the target area of prostate cancer. The anatomical relationship of tissues in the pelvic cavity is complicated, and there are OAR include bladder and rectum. We compared the effect of prostate target area of CT, T1WI and T2WI sequence on the dosimetry of bladder and rectum. Compared with CT group, the exposure doses of bladder and rectum (when exposure dose < 30Gy) of MRI group decreased, but there was no statistical difference, which consistent with the research of OT et al (16).
Previous studies have con rmed that mpMRI has a better resolution of the internal anatomical structure of the prostate. These imaging methods can accurately detect the presence of malignancy from normal tissue and the extent of tumor invasion clearly, which improves the accuracy of prostate cancer biological target area and optimizes radiotherapy planning greatly. Although mpMRI is superior to CT in determining the biological target area, considering the geometric deformation possibility during process, the MRI-only target delineation based on has a certain risk of target edge missing, which needs to be combined with pathological slices for further research.

Conclusions
Target area delineation is a crucial link in precision radiotherapy progress. Accurate target delineation contributes to improve the quality of planning and reduce the radiation level of healthy tissues, thereby improving patient prognosis. This study indicated that in prostate cancer radiotherapy, the target area based on mpMRI (T1WI, T2WI) is more accurate than CT, which can signi cantly reduce the target area and the volume of the rectum and bladder exposed to high radiation dose, and improve the positioning capability. Since T2WI is superior in target outlining, it is worth prioritizing to application. This study was approved by the ethics committee of Shandong Cancer Hospital. All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Written informed consent was obtained from individual participants.

Consent for publication
All data published here are under the consent for publication. Written informed consent was obtained from all individual participants included in the study.

Availability of data and materials
The datasets generated and analyzed during the present study are available from the corresponding author on reasonable request.

Figure 1
The relationship of CTV de ned by T1WI, T2WI and CT imaging   The spatial distributions maps of the PTV de ned by CT, T1WI and T2WI images (a) The targets outlined by CT, T1WI and T2WI images are roughly the same. (b) T2WI provide signi cantly more speci c target outlines than CT. The discrepancy in delineation between the two groups was mainly in the basal and apical areas of the prostate. There was a small discrepancy between the target regions de ned by T1WI and T2WI. The sectional imaging of CT (a), T1WI (b) and T2WI (c) of the same patient along the same plane of view. (a).the boundary between prostate gland and surrounding tissue was unclear. (b).the boundary between the visible gland and peri-glandular adipose tissue, levator ani muscle and the posterior rectum could be seen on T1WI image. (c). T2WI image clearly showed the prostatic capsule, which was in strong contrast with peripheral zone of the prostate (peri-prostate tissue, mainly composed of mesentery, neurovascular bundle and levator ani muscle).