Abstract
Background/Aim: Cone-beam computed tomography (CBCT) is the most commonly used system in modern radiotherapy of prostate cancer for daily positioning verification. The use of intraprostatic radiopaque fiducials (FMs) may be added to CBCT. We wanted to investigate the possible advantage of using FMs in daily CBCT repositioning. Materials and Methods: We selected three CBCTs for each treatment course for 13 patients (seven with and six without use of FMs) treated at our centre. Seven experienced Radiation Oncologists retrospectively reviewed the CBCTs, recording couch movements for correct patient positioning, and time spent to do it. Analysis of variance and t-test were carried out for comparison of different groups and for differences in mean values of the movements recorded (with p<0.05 as significance level). Results: No statistically significant difference was found between operators in the analysis of images with FMs nor of images without them. A difference was only found in the mean corrections in couch rotation and pitch angle, which were higher in the FM group, and in the mean time for image analysis, which was shorter in this group. Using the van Herk formula, we found a possible reduction of clinical target volume and planning target volume margins for the FM group. Conclusion: According to our study, the use of intraprostatic FMs in daily CBCT seems useful for better detection of and correction for non-negligible rotational errors. Furthermore, FMs reduced the time to treatment start, which is very important in reducing the risk of intrafraction organ motion. These results need to be confirmed by further studies.
Prostate cancer is the most common cancer in men, with 268,490 estimated new cases in the United States in the year 2022 (1). Radiotherapy (RT), as well as surgery, is one of the main curative treatment modalities.
In RT of prostate cancer, target intra- and interfractional motion is a well-known issue; it is mainly related to the different daily filling of the bladder and rectum and to the bowel peristalsis, which displace the prostate.
Various imaging systems are routinely used for daily image-guided radiotherapy (IGRT) to avoid missing the therapy target. Positioning control for more precise treatments may allow for reduction of margins, delivery of larger doses at smaller planned target volumes, and reduced toxicity of adjacent tissues. Gill et al., in a series of patients with prostate cancer who were treated radically with 78 Gy with IGRT, reported less severe urinary frequency, diarrhoea and fatigue during treatment compared to patients receiving 74 Gy without IGRT, using the same margins (2). IGRT showed an improvement in biochemical tumour control among high-risk patients and a lower rate of late urinary toxicity compared with high-dose non-image-guided intensity-modulated RT (3).
Daily orthogonal kilovoltage or megavoltage x-ray imaging using intraprostatic implanted fiducial markers (FMs), and cone-beam computed tomography (CBCT), are two of the most common systems for assessing interfractional target motion and to correct daily setup (4). Images generated at the time of treatment are compared with previously obtained planar or three-dimensional images; an isocentre match is made, and final patient’ couch shifts are performed before treatment delivery.
The usefulness of using the CBCT plus FMs is not clear; there are few comparative studies, and with conflicting results (5, 6). The implantation of FMs is an invasive procedure and involves a non-negligible risk of complications (7), for this reason many centres prefer to use CBCT alone.
The duration of treatment sessions is a critical topic for the amount of intrafraction prostatic motion (8-11): the effects of intrafractional motion on prostate treatment are significantly smaller and the required margins can be therefore reduced when the treatment is shortened.
The aim of our study was to compare the geometric interobserver differences and the time consumed in patients with and without intraprostatic FMs in the daily assessment of CBCT for prostate IGRT.
Materials and Methods
At our centre, RT in intermediate-risk prostate cancer, according to medical evaluation and the will of the patient, is performed with IGRT through CBCT with or without prostatic FMs, and with two possible schemes: 60 or 70 Gy in 20 or 28 daily sessions, respectively. Three intraprostatic radiopaque FMs are placed transperineally at least 7 days before planning CT. Patients are instructed to perform a rectal enema before simulation CT, and to arrive with an empty bowel and full bladder for simulation CT and treatment. With the patient in the supine position, a CT scan is performed from top of the L3 vertebral body to below the lesser trochanters. CT simulation is performed using a GE Hi-Speed NX (GE Healthcare Italy, Milan, Italy). Reference CTs are acquired using a 120-140 kV beam and 3.0-mm slice thickness and 1-mm into the FM region identified on the initial anterior–posterior (AP) scout acquisition. Anterior, right and left lateral permanent skin markers are placed. The target includes the prostate gland and seminal vesicles. Plans are optimized for volumetric-modulated arc therapy, generally with two full arcs. Progressive resolution optimization is used to modulate multileaf collimator shape and beam intensity during the gantry rotation. Delivery is through a 6-MV beam from a linear accelerator, TrueBeam v.2.7 (Varian Medical Systems, Palo Alto, CA, USA), equipped with a Millenium MLC-120. Planning computing is performed in the Eclipse RT planning systems (Aria v.15.6; Varian Medical Systems) which employs an anisotropic analytical algorithm for photon-beam calculation with a grid size of 2.5 mm. Daily CBCTs are acquired using an On-Board Imager (Varian Medical Systems), with a source voltage of 140 kVp and a source-to-detector distance of 150 cm. The CBCT acquisition protocol uses a half fan filter arrangement with a 360° gantry rotation. Reconstruction is performed with a 512×512 resolution and 2.0 mm slice thickness.
Before daily treatment, patients are positioned according to the skin markers using three lasers in the treatment room, a CBCT is performed for all treatment sessions and evaluated by a Radiation Oncologist; after an automatic CTV- or FM-based co-registration of CBCT and CT simulation, final corrections are manually performed. The translations of the treatment couch are corrected for displacement of the target with six possible degrees of freedom of the patient table before treatment delivery.
For this study, 13 consecutive patients treated radically for localized prostatic cancer from January to May 2022 were identified; six patients (46%) had no intraprostatic implanted FMs, while seven (54%) had them. The median age in the group without FMs was 75 (range=67-84) years and was 76 (range=69-79) years in the group with them. According to the eighth edition of the American Joint Committee on Cancer Cancer Staging Manual (12), in the group without FMs, two patients had IIB stage and four had stage IIC cancer; in the FM group, five patients had IIB stage cancer, one patient stage IIC, and one patient stage IVB due to a secondary sacral bony localization of disease.
One patient (16.7%) of the group without FMs had 20 treatment sessions, five patients (83.3%) had 28 sessions. In the FM group, three patients (42.9%) were treated in 20 sessions and four (57.1%) in 28.
Three CBCT examinations were selected for each treatment course: At the first session, at the last, and one in the middle; the results of a total of 39 examinations were collected.
Seven Radiation Oncologists, with at least 7 years of experience in CBCT evaluation, retrospectively reviewed the selected 39 examination in June 2022. They recorded the movements of the treatment couch to place the patient in the correct position with respect to the planning CT, and the time needed to do it, to identify possible differences in the two settings of patients.
Analysis of variance was carried out for comparison of different groups to assess inter- and intra-observer variation; t-test for unpaired data was used to test differences in mean values of different parameters related to movement. We tested mean couch shifts in the left–right (LR), anterior–posterior (AP) and superior–inferior (SI) directions and mean values of couch rotations (X°, Y° and Z° for roll, rotation, and pitch, respectively). We also compared the mean time spent for each CBCT evaluation and CTV-planning target volume margins derived with van Herk formula (13).
For all the statistical tests, p<0.05 was set as the significance level; analysis was carried out with Microsoft Excel (2013, Microsoft Corporation, Redmond, WA, USA).
Results
In the analysis of variance, no statistically significant difference was found between operators nor in the analysis of images with FMs nor in images without them.
The mean couch shifts in LR, AP and SI directions, as well as mean values for couch rotations for CBCT with and without FMs are summarized in Table I. The mean time spent for CBCT evaluation was 126±46 s for CBCT without FMs and 109±33 s for CBCT with FMs, a statistically significant difference (p=0.03).
Mean couch shifts and rotation in cone-beam computed tomography with and without fiducial markers (FMs). LR: Left–right; AP: anterior–posterior; SI: superior–inferior.
A statistically significant difference between the two groups of patients was also found in the mean values of angle correction for rotation and pitch, being higher in the group with FMs (Y°, p=0.04 and Z°, p=0.01), and in mean time for image analysis, being shorter in the same group (p=0.03).
Using the van Herk formula (13) we derived CTV-planning target volume margins in the three directions, obtaining 8 mm in the LR direction, 6 mm in the AP direction and 13 mm in the SI for CBCT without FMs versus 5 mm, 3 mm and 9 mm, respectively, with FMs.
Discussion
The implantation of FMs is an invasive procedure. In their review, De Bari et al. reported of 3.8-15% incidence of haematuria, 18.5% of haematospermia and 4-9.1% of rectal bleeding (7). High incidences of infectious complications had also been reported. Loh et al. in their series of transrectal prostatic FM implantation reported a urinary infection rate of 7.7%, with a total of 2.8% of patients requiring hospital admission for infective complications (14). For this reason, many RT centres prefer to use a non-invasive approach for prostate IGRT such as CBCT.
In reality, the risk of complications is high if intraprostatic landmarks are implanted transrectally but very low if implanted transperineally. The largest series of transperineal marker implants was performed by Saad et al., who recorded two cases of urinary infection and two cases of transient gross haematuria in 579 patients overall (15). Our experience in two series, previously reported, also confirmed the safety of this procedure, with about 1-1.5% of transient, self-limiting gross haematuria in 101 and 133 patients, respectively, and no other complication (16, 17).
There are very few studies assessing the utility of FMs in RT of prostate cancer with CBCT for daily patient repositioning. Deegan et al. quantified inter-observer variability and compared CBCT localisation when Radiation Oncologists aligned FMs or soft tissue on CBCT; inter-observer variability was considerably greater for soft-tissue alignment compared with the use of FMs, and large differences between soft-tissue localisation and FM position were found (5).
Yildirim et al. recorded organ motion and set-up errors during RT treatment using the same CBCTs, first with soft-tissue- and then FM-based corrections, without the use of rotations. Rotational corrections were considered to be insignificant and therefore ignored in this study. The authors found that the prostate displacements observed in CBCT corrections based on prostate gland and FMs were very similar and FMs did not provide any significant contribution (6).
In our opinion, in the series of Yildirim et al. (6), the soft-tissue-based CBCT matching with the reference simulation may have been influenced by the presence of the FMs in the same patients and visible in this evaluation; furthermore, it is not clear how the authors were able to define rotational displacements as being insignificant, not having recorded them.
According to our study, there was no difference between operators in the analysis of images. This is probably due to the expertise of the seven Radiation Oncologists. The mean couch shifts in all directions were smaller for CBCT with FMs even though with no statistical significance. For this reason, margins calculated with the van Herk formula (13) were also smaller.
The use of intraprostatic FMs in daily CBCT seems useful for better detection of the presence of rotational errors [especially for rotation (p=0.04) and pitch (p=0.01)] allowing for complete positioning correction; in our experience, rotational errors play an important role in correct positioning and cannot be ignored.
Furthermore, reducing the time spent on analysing the CBCT prior to treatment speeds up the start of treatment, which is very important to reduce intrafractional uncertainty, which is time-dependent (8-11); as the duration of an RT session increases, the displacements of the prostate increase, consequently the margins must be increased (10).
To our knowledge, no previous study has ever compared the time taken by Radiation Oncologists in the daily pre-treatment assessment of CT with or without FMs. The use of prostate FMs, implanted transperineally, even if an invasive procedure, in our opinion could be useful for better prostate targeting and to reduce the risk of intrafractional organ motion. This might justify a reduction in margins in IGRT with FMs compared to IGRT without, although further studies are recommended.
Footnotes
Authors’ Contributions
Conceptualization: Luigi. De Cicco Investigation: Luigi De Cicco, Rossella Margherita Mancuso, Elena Petazzi, Angelo Giovanni Lanceni, Elisa Della Bosca, Sandra Buttignol and Antonio Starace. Data analysis: Luca Marzoli. Project administration: Luigi De Cicco and Rossella Margherita Mancuso. Writing original draft: Luigi De Cicco, Luca Marzoli and Rita Lorusso. Review and editing: Barbara Bortolato and Claudio Verusio.
Conflicts of Interest
The Authors declare no conflicts of interest in relation to this study.
- Received November 18, 2022.
- Revision received November 30, 2022.
- Accepted December 5, 2022.
- Copyright © 2023 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
