Biomedical advances and future prospects of high-precision three-dimensional radiotherapy and four-dimensional radiotherapy

Hiroki SHIRATO

doi:10.2183/pjab.99.024

Abstract

Biomedical advances of external-beam radiotherapy (EBRT) with improvements in physical accuracy are reviewed. High-precision (±1 mm) three-dimensional radiotherapy (3DRT) can utilize respective therapeutic open doors in the tumor control probability curve and in the normal tissue complication probability curve instead of the one single therapeutic window in two-dimensional EBRT. High-precision 3DRT achieved higher tumor control and probable survival rates for patients with small peripheral lung and liver cancers. Four-dimensional radiotherapy (4DRT), which can reduce uncertainties in 3DRT due to organ motion by real-time (every 0.1–1 s) tumor-tracking and immediate (0.1–1 s) irradiation, have achieved reduced adverse effects for prostate and pancreatic tumors near the digestive tract and with similar or better tumor control. Particle beam therapy improved tumor control and probable survival for patients with large liver tumors. The clinical outcomes of locally advanced or multiple tumors located near serial-type organs can theoretically be improved further by integrating the 4DRT concept with particle beams.

1. Introduction

External-beam radiotherapy (EBRT) was started as an application of physics and engineering to the cancer treatment at the beginning of the 20th century, and developed to become a standard curative treatment for many types of cancer.¹⁾^–³⁾ Without surgical or pharmaceutical intervention, cancer cells in the tumor are killed after a certain amount of exposure to radiation from outside of the body in EBRT. Since it has been known that ionizing radiation also affects normal cells around a tumor, how to concentrate the absorbed dose to the tumor while protecting surrounding normal tissues became the main issue in the history of EBRT.³⁾^–⁵⁾ During the period when X-ray fluoroscopy was used for treatment planning of EBRT, only two-dimensional (2D) information based on the bone structure and contrast agent was available.⁴⁾ With the invention of computed tomography (CT),⁶⁾ three-dimensional (3D) co-ordinates and the electron density of specific anatomical structures became available for dose calculations with an accuracy of 1 mm in the 1970s⁷⁾ and became used in computerized 3D treatment planning in the 1980s.⁸⁾^,⁹⁾ Defining the treatment-related volumes such as clinical target volume (CTV) and organs at risk (OAR) by the International Commission on Radiation Units and Measurements (ICRU) Report 50¹⁰⁾ and 62¹¹⁾ has revolutionized the concept of EBRT. Distribution of the absorbed dose in the body has been further improved using inverse planning and intensity modulation radiotherapy (IMRT) since the 1990s.¹²⁾^,¹³⁾ More recently, four-dimensional (4D) information, i.e., 3D spatial coordinates at different time points, became available for treatment planning using CT scan and magnetic resonance imaging (MRI).¹⁴⁾^,¹⁵⁾ Particle beams such as proton and carbon ion beams, which are physically more suitable than photons beams such as X-rays and gamma-rays to focus ionization to the tumor, have emerged as new EBRT modalities.¹⁶⁾^,¹⁷⁾ With the progress in the treatment planning and therapeutic machinery, highly precise positioning of tumors in spatial-temporal co-ordinates became a dominating issue.¹⁶⁾^,¹⁸⁾^–²⁰⁾ However, it has not been established whether we could increase cure rates without increases in radiation-induced adverse effects in cancer patients with the improvement in physical accuracy. In this article, biomedical advances accompanying the improvement in the physical accuracy with recent EBRT applications are reviewed.

In the two-dimensional radiotherapy (2DRT) era, multiple fractionation of absorbed doses (J/Kg = Gy) in a specific period of time was the primary strategy to reduce serious late adverse effects without deterioration in the tumor control rate.²¹⁾ In the 2DRT era, the absorbed dose to the cancer was equal to the normal tissue near the tumor due to the lack of 3D information (Fig. 1(a)). The classical guiding principle of 2DRT was to use the one “therapeutic window” between the dose-response curves for the tumor and the surrounding normal tissue just like in pharmaceutical therapy (Fig. 1(b)).²²⁾ If a tumor is more radiosensitive than the surrounding normal tissue, we can prescribe the dose to the tumor between the threshold of the tumor control probability (TCP) (for example 95%) and that of the normal tissue complication probability (NTCP) (for example 5%). The ratio of the TCP for a threshold level of NTCP has been called the therapeutic ratio. In 2DRT, if the normal tissue near the tumor is more radiosensitive, there is no therapeutic window (the no therapeutic window problem). For example, if a cancer has developed behind the stomach, which is more radiosensitive than common cancers (Fig. 1(c)), there was no therapeutic window in the 2DRT era (Fig. 1(d)). However, using three-dimensional radiotherapy (3DRT), the radiation dose distribution within the body can be controlled intentionally to reduce the dose to the normal tissue, the stomach in the example, near the tumor (Fig. 1(e)). We may say that the guiding principle of 3DRT became to use prescriptible dose provided by the “respective therapeutic open doors in the TCP and NTCP curves” which can be illustrated by using a bird’s eye view replacing the classical therapeutic window in 2DRT as suggested in Fig. 1(f).

Fig. 1.

(a) A skin cancer (squamous cell carcinoma) before and 2-years after 2DRT. (b) The classical therapeutic window (dotted rectangle) in 2DRT between the dose-response curve for TCP (red) and that for NTCP (blue) for the patient shown in (a). “Brick walls” represent the dose range that is not good for prescription. (c) Commonly occurring dose distributions of 2DRT for tumors behind the stomach on a CT slice. (d) Dose-response curve for TCP and that for NTCP for the patient shown in (c) in 2DRT. There is no therapeutic window in 2DRT since the dose-response curve for TCP (red) is located at the higher dose side of the NTCP (blue). (e) Common dose distribution of 3DRT for the same patient showing reductions in the dose to the stomach. (f) Dose-response curve for TCP and for NTCP in different dose-response co-ordinates for the patient shown in (e) in 3DRT. There are therapeutic open doors in 3DRT for the tumor shown in (e) since doses higher than the threshold of TCP (95%) is prescriptible for the tumor with the reduced dose less than the threshold of NTCP (5%) for the stomach in 3DRT.

With 3DRT, it becomes possible to solve the no therapeutic window problem in 2DRT. We become able to give a curative and sufficient dose to the tumor using 3DRT as long as the normal tissue can be anatomically distinguished from the tumor. With the prevalence of 3DRT, positioning accuracy became one of the performance effects that distinguish 3DRT. Speaking generally, the 3DRT can be classified into conventional 3DRT and high-precision 3DRT both for photon and particle beam therapy. When 3DRT fulfills the following two requirements, it is called high-precision 3DRT in this review. The first requirement is a planning and positioning technology that precisely captures the spatial coordinates of the patient tumor and its surrounding normal tissues at the start of irradiation with an accuracy of ±1 mm. The second requirement is the precisely focused delivery of the radiation beam to the static target with an accuracy of ±1 mm. For example, image-guided radiotherapy (IGRT), which utilizes the 3D imaging modality for the daily set-up of patients²³⁾^–²⁵⁾; and stereotactic radiotherapy (SRT), which utilizes a fixation device for immobilization of the patient²⁶⁾ with recently obtained 3D images, are regarded as high-precision 3DRT when used with a precisely focused delivery system of the radiation beam to the static target providing an accuracy of ±1 mm.

However, when the tumor is moving with internal organ motion such as respiration or bowel movement, we need to add a margin around the tumor and irradiate a larger volume than the actual tumor even with high-precision 3DRT (Fig. 2(a)–(d)).²⁷⁾^,²⁸⁾ When there are radiosensitive organs in the margin around the tumor, the dose to the normal tissues increases to become similar to that of 2DRT. When the dose to the normal tissue cannot be reduced below the threshold of the NTCP without lowering the dose to the tumor (and thus the TCP), “we cannot enter the therapeutic open doors because of organ motion” (the organ motion problem) and we can use only an insufficient palliative dose to the cancer even with high-precision 3DRT.

Fig. 2.

Illustration of reductions in volume irradiated in 4DRT compared with high-precision 3DRT for a small lung cancer in the right lung.¹⁴⁾ (a) The orange line represents a tumor in the coronal CT image during normal breathing without holding the breath. The white circle represents 90% and 80% isodose curves in high-precision 3DRT. (b) The dose distribution superimposed on the CT at the end of inhaling and (c) at exhaling. A part of cancer was outside the 80% isodose curve both at the end of inhaling and exhaling. (d) The white oval line represents the area to be covered by the 80–90% isodose curve in high-precision 3DRT. (e) Example of dose distribution in a moving lung phantom, which is moving at 12 cycles per minute, using high-precision 3DRT. X-ray film was installed in the lung phantom. (f) The dose distribution in the same moving lung phantom using 4DRT.³⁰⁾ (Reproduced from Int. J. Radiat. Oncol. Biol. Phys. 48, 435–442 ((a)–(d)) and Lancet 353, 1331–1332 ((e), (f)).)

2. Four-dimensional radiation therapy

To solve the organ motion problem in high-precision 3DRT, time-resolution in the treatment planning and dose delivery was required to be improved further. Four-dimensional, time and space, treatment planning was started by integrating the respiratory phases and 3D movements of tumors using CT and MRI in the late 1990s.¹⁴⁾^,¹⁵⁾^,²⁰⁾^,²⁷⁾^,²⁹⁾ Following the 4D treatment planning, a new EBRT modality with an attempt to improve the time resolution in the dose delivery system in high-precision 3DRT was introduced and termed 4DRT.¹⁴⁾^,²⁰⁾^,³⁰⁾^–³³⁾ When high-precision 3DRT fulfills the following two requirements it is called 4DRT in this review. The first requirement is the technology to acquire the 3D position of the tumor in real-time, using real-time tumor tracking technology. The second requirement is for the technology to provide immediately focused irradiation to the moving target. The definition of “real-time” and “immediate” will be discussed further below. The 4DRT application has been developed to narrow the margin required around the cancer by reducing the blurring of the dose distribution due to organ motion (Fig. 2(e), (f)). We can provide a sufficient dose to a moving tumor, reducing the dose to the surrounding normal tissue with real-time tracking of tumor and immediately focused irradiation technology such as beam-gating or dynamic tracking of the therapeutic beam adjusted to the tumor motion. We may express this as that the guiding principle of 4DRT is to “Enter the therapeutic open doors by reducing the margin for organ motion”. With 4DRT, we can solve the organ motion problem in high-precision 3DRT and give a curative and sufficient dose to the moving tumor.

In ideal 4DRT, the 4D coordinates (space ($\vec{A}$), time (t)) of whole parts of the human body at risk should be used in the dose calculations and the treatment beam should be focused to the tumor with the same precision as with high-precision 3DRT even when the tumor is moving in concert with the organ motion. An ideal 4DRT has not been developed but this review will summarize how development has progressed so far, with the aim of perfecting it further.

2.1. The original 4DRT system.

To solve the organ motion problem in the 2DRT era, gated EBRT, which was comprised of 2DRT, 1-dimensional monitoring of the respiratory motion, and immediate focused irradiation technology, was developed.³⁴⁾^,³⁵⁾ Fiducial markers are suggested for use as a surrogate of the internal position of the tumor in 3DRT.³⁶⁾ During this, real-time tracking technology had been developed in several industries using a method for searching and finding the location of a template image in a larger image in real-time. Combining real-time tracking technology, internal fiducial markers, 3D calculations of the tumor position, and gated 3DRT technology, resulted in a new EBRT system becoming developed, this was named real-time tumor-tracking radiotherapy (RTRT), and started to be used clinically in 1999.³⁰⁾ Details of the RTRT is shown in Fig. 3. Several authors pointed-out that RTRT is the initial real-time 3D IGRT since it used real-time tumor-tracking technology during high-precision 3DRT.³⁷⁾^,³⁸⁾ Because RTRT used immediately focused irradiation to the moving tumor as well, this system is regarded as the initial 4DRT system according to the definition presented above. The RTRT is also the initial real-time-image gated radiotherapy (RGRT), which combines real-time tumor-tracking technology with beam gating technology in general.

Fig. 3.

(a) Gold markers of 1.5–2.0 mm diameter inserted near a lung tumor using a fiberscope under fluoroscopy by pulmonologists before CT scanning for planning. Physical relationship between the marker position and the volumetric details (x,y,z) of the cancer is obtained by CT in the treatment planning prior to the delivery of the therapeutic X-ray beam.⁴⁷⁾ (b) Real-time-tracking radiotherapy (RTRT) system.⁴⁸⁾ (c) Two actual fluoroscopic images are displayed on the computer console, which is in the control room of the treatment, of the RTRT system. The 3D co-ordinates of the marker in the patient were automatically calculated every 0.033 s by real-time tracking of the marker using pattern matching of the marker to the template image.³¹⁾ (d) Two out of 4 sets of fluoroscopes around the medical linear accelerator in the room used during treatment. In a static phantom, the localization accuracy was 0.24 ± 0.34 mm throughout the fluoroscopic field.³¹⁾ (e) With accurate knowledge of the position of the fiducial marker near the moving tumor, the patient is irradiated with a therapeutic X-ray beam only while the marker is positioned in a gating window. (f) The planned 3D coordinates of the marker are transferred from the CT to the RTRT system prior to the delivery of the therapeutic X-ray beam, and the projected position of the marker is superimposed on each of the corresponding fluoroscopic images. The actual position of the marker can be seen in the fluoroscopic images. When the actual position is inside the gating window (±2 mm), the treatment beam is on. When the actual position of the marker is outside of the gating window, the treatment beam is off.¹⁴⁾ (g) Range (mm) of movement for left-right (x), cranio-caudal (y), and antero-posterior (z) during the beam-off period, which represents the high-precision 3DRT (left) and beam-on period, which represents the 4DRT (right), in 4 patients in RTRT under natural breathing.³⁹⁾ (Reproduced from Cancer 95, 1720–1727 (a), Int. J. Radiat. Oncol. Biol. Phys. 48, 435–442 ((b), (f)), 1187–1195 (c), 1591–1597 (d), and 51, 304–310 (g).)

As an example of the advantage of the 4DRT system, the 3D motion amplitudes of lung cancer during radiotherapy could be reduced significantly compared to high-precision 3DRT. In 4 patients with peripheral lung cancer, the amplitudes (99th percentile) of cancer movement was significantly reduced from 5.5–10.0 mm in the left-right (LR), 6.8–15.9 mm in the cranio-caudal (CC), and 8.1–14.6 mm in the antero-posterior (AP) directions during beam-off periods, a performance which is equivalent to high-precision 3DRT, to within 5.3 mm in all directions during the beam-on period in RTRT (Fig. 3(g)).³⁹⁾ The output duty cycle of the linear accelerator (linac) ranged from 26.7% to 56.7% for the 4 lung patients, sufficiently large to be used in usual cancer centers.

The improvement in dose distribution in 4DRT can be demonstrated with film measurements for moving phantoms. The dose distribution of IMRT for moving phantoms using RTRT was much better than that for moving phantoms using high-precision 3DRT and similar to that for static phantoms using high-precision 3DRT (Fig. 4(a)). Engelsman et al. have investigated how much it is possible to reduce unnecessary margins by 4DRT for lung and liver cancers using the patient data obtained from the RTRT system (Fig. 4(b)).⁴⁰⁾ They showed that the required margin is a nonlinear function of the amplitudes of the cancer movement. Only a small reduction in required margin was observed for the amplitudes of cancer movement up to 10 mm. However, for larger movements, 4DRT can allow a substantial reduction in margins of about 7 or 12 mm depending on the dose gradient prior to blurring. The results of Engelsman et al. indicated that 4DRT can reduce the margin and thus the irradiated volume to normal tissue, and increase the minimum dose at the edge of the moving tumor compared to high-precision 3DRT.

Fig. 4.

(a) Dose distribution of IMRT measured with X-ray film in (left) high-precision 3DRT for a static phantom, (center) high-precision 3DRT for a moving phantom, and (right) RTRT with a ±2 mm gating window for a moving phantom, all using the same treatment plan. The high dose region (red) is smaller and the low dose region (green and blue) is wider in high-precision 3DRT for the moving phantom. The dose distribution of RTRT for a moving phantom was similar to the 3DRT for a static phantom. (b) Dose profiles for an 8MV X-ray penumbra in the lungs. The grey line is the unblurred profile which is either blurred with a breathing probability density function (PDF) of the tumor position (broken black line) or with a Gaussian PDF with a standard deviation of the breathing PDF (solid black line). The inset shows the specific breathing PDF.⁴⁰⁾ (Reproduced from Phys. Med. Biol. 50, 477–490 (b).)

2.2. Prediction modeling for internal tumor motion.

Various international collaborative studies about tumor motion and prediction models were conducted using patient data stored in the 4DRT system in the 2000s. Sharp et al. have investigated how prediction models such as linear prediction, neural network prediction, and Kalman filtering can reduce the requirements of the imaging rate.⁴¹⁾ The real question that needed to be addressed was how much we can increase the system latency, from the detection of the tumor position to the delivery of the therapeutic beam, by the prediction models. This showed that if the real-time tumor tracking technology had a latency of 0.2 s or greater, and for technology that has imaging rates of 10 Hz or slower, these prediction models are useful.⁴¹⁾ Wu et al. have presented a one-dimensional version of a finite state model (Fig. 5) to characterize a wide variety of patient breathing patterns.⁴²⁾ They found that: 1) breathing patterns are patient specific, 2) generally, the end-of-exhalation (EOE) state has the longest duration, while the inhalation (IN) and exhalation (EX) durations are shorter than EOE, 3) the IN state has the longest travel distance and EOE has the shortest distance, 4) the EOE velocity is small and has very limited variation, while both IN and EX velocities vary widely, 5) there is a strong linear correlation between the velocity of the tumor(marker) and travel distance of the tumor(marker) when averaged by patient, 6) generally, the IN→EX motion correlations are better than the EX→IN correlations, which means that the EX motion properties will follow their immediately previous IN properties, but not the opposite, and 7) motion characteristics are associated with the beginning state of the breathing cycle.

Fig. 5.

(a) A finite state model for the cranio-caudal (CC) tumor motion. EX: exhaled status, EOE: end-of-exhaled status, IN: inhaled status. (b) Distance changing patterns for a specific patient. (c) Histogram of distance changes for the same patient.⁴²⁾ (Reproduced from Phys. Med. Biol. 52, 4761–4774 ((a)–(c)).)

Berbeco et al. and his colleagues have estimated the residual motion of lung tumors if radiotherapy were to be gated only to the external respiratory surrogates without imaging of the internal marker and without prediction models.⁴³⁾ Using a modified RTRT system which can acquire the signals from external respiratory surrogates and internal fiducial markers simultaneously, they found that the residual motion (95th percentile) was 0.7–5.8, 0.8–6.0, and 0.9–6.2 mm for 20%, 30%, and 40% duty cycle windows respectively when only external respiratory surrogates were used. They have also noted large fluctuations (>300%) in the occasional residual motion between some beams. Overall, the mean beam-to-beam variation was as large as 42% from the previous treatment beam and the day-to-day variation was 34% from the previous day for phase-based gating. Using the same RTRT dataset, Kanoulas et al. have confirmed that the relationship between external respiratory surrogates and internal fiducial markers changes with time and that large errors in prediction occurs if only the external surrogate is used to predict the internal tumor position (Fig. 6(a)).⁴⁴⁾ To derive the tumor position from an external surrogate to predict the internal tumor position during treatment, periodically updating methods for internal/external correlations using occasional dual X-ray radiographic imaging and updating of the linear correlation function were shown to be useful.⁴⁴⁾ Ionascu et al., however, have shown that there can be significant and unpredictable time shifts and amplitude mismatches between internal fiducial marker and external surrogates when only a linear correlation function is used.⁴⁵⁾ These effects could be caused by the 0.4–0.6 s instantaneous time shifts that induce amplitude mismatches in the range of 2.5–4.7 mm. Seppenwoolde et al. have shown that the addition of a polynominal model is useful to reduce the residual error for the tumors which showed internal hysteresis or had a time delay between the internal and external motion (Fig. 6(b)).⁴⁶⁾

Fig. 6.

(a) External surrogate position versus tumor position in the cranio-caudal (CC) direction during patient setup and treatment sessions. The line that best fits the training data along with the constantly obtained external surrogate position is used to predict the tumor position during treatment. The dataset is obtained from a patient with a peripheral lung tumor at the lower lung.⁴⁴⁾ (b) 3D representation of the treatment error for irradiation with a polynomial prediction (dots) and without using the prediction (solid line). The white dot represents the predicted tumor position. Polynomial model reduced the residual error compared with the linear model when there is hysteresis.⁴⁶⁾ (Reproduced from Phys. Med. Biol. 52, 5443–5456 (a) and Med. Phys. 34, 2774–2784 (b).)

2.3. New developments in 4DRT technology.

There have been many developments since 2000 attempting to simplify or improve the shortcomings of the initial 4DRT system such as increases in X-ray exposure due to fluoroscopy, prolonged treatment time due to gating, fiducial marking,⁴⁷⁾^,⁴⁸⁾ absence of volumetric tracking, and non-compatibility with particle beam therapy such as proton beam therapy (PBT) and carbon beam therapy (CBT). Equipment for 4DRT which has obtained approval from regulatory agencies such as the Pharmaceuticals and Medical Device agency of Japan (PMDA), the Food and Drug Administration of U.S.A. (FDA), and the Medical Device Directive (MDD) are shown in Fig. 7.

Fig. 7.

Examples of 4DRT systems which obtained approval from regulatory agencies. (a) SyncTraX^® connected to the linear accelerator with a triode electron gun (provided by Shimadzu), (b) Respiratory Tracking System^® with CyberKnife^® (provided by Accuray), (c) Oxray^® (provided by Hitachi), (d) MRIdian^® (provided by ViewRay), (e) Unity^® (provided by Elekta), (f) PROBEAT-RT^® (provided by Hitachi).

2.3.1. Reduction of exposure with fluoroscopy.

Pulsed fluoroscopy at 30 Hz was used in the original RTRT system and was more suitable than continuous fluoroscopy for real-time tumor tracking since the irradiation dose due to exposure being lower with pulsed fluoroscopy.⁴⁹⁾ The lower frequency of pulsed exposure, such as 1, 5, 10, and 15 Hz, with motion prediction models became available to reduce X-ray exposure further in the second- and third-versions of the RTRT linac systems (Mitsubishi Electronics, Tokyo, Japan) or SyncTraX^® (Shimadzu, Kyoto, Japan) connected to a triode-type linac (Varian Medical System, California, U.S.) (Fig. 7(a)).⁵⁰⁾^,⁵¹⁾ At the time of writing of this review (July 2023), a total of 35 RTRT systems have been installed in Japanese hospitals (personal communication). Four-dimensional set-up based on the daily trajectory of the internal fiducial marker at the start of treatment followed by reduced rates of exposure during the delivery of the therapeutic beam, when appropriate, became the standard procedure in the hospitals where new RTRT systems are installed.⁵²⁾

Combinations of external markers on the skin, internal fiducial markers, and dual radiography using an X-ray generator with a lower heat unit, and prediction models for internal/external correlation have been successfully commercialized in several systems.⁴⁶⁾^,⁵³⁾^–⁵⁷⁾ Respiratory Tracking System^® (RTS) on CyberKnife^® (Accuray Inc., California, U.S.) (Fig. 7(b)) does not allow continuous acquisition of X-ray images like the RTRT system and the tumor position is predicted from external respiratory surrogates with periodically updated internal/external correlations.⁴⁶⁾ The 3D internal tumor position is determined at discrete time points by automatic detection of the internal fiducial markers in two orthogonal X-ray images: simultaneously, the positions of the external markers are measured in RTS. During the treatment, the relationship between internal and external marker positions is continuously accounted for by the X-ray images and is regularly checked and updated. In a simulation study, treatment errors due to breathing motion were reduced by using RTS in all 8 patients studied.⁴⁶⁾ The Synchrony^® on Radixact^® (Accuray Inc., California, U.S.) uses sequential radiographs from different angles, combined with external markers on the skin, and prediction models.⁵⁷⁾ Oxray^® (Hitachi, Tokyo, Japan) which is the successor model of Vero 4DRT^®/MHI-TM2000^® (BrainLab, Munich, Germany/Mitsubishi Heavy Industry, Tokyo, Japan) (Fig. 7(c)) also uses an updated internal/external correlation model and periodic fluoroscopic observations.⁵⁸⁾ The three-dimensional probability density function for the target position has been investigated to compensate for the drawback of real-time 2D-IGRT and is used for prostate cancer treatment.³⁷⁾^,⁵⁹⁾ Electromagnetic transponders (EMT) such as Calypso^® (Varian Medical System, California, U.S.) are implanted near the tumor and used for real-time tumor tracking in various moving organs including the prostate and lung cancers with or without image-guidance.⁶⁰⁾^–⁶²⁾ Ultrasound systems have also been used for solid organs such as the prostate gland and liver.⁶³⁾^,⁶⁴⁾ Further, respiratory gating without real-time internal imaging during 3DRT has also been used as an advanced high-precision 3DRT or a simplified 4DRT.⁶⁵⁾^–⁶⁷⁾

2.3.2. Reductions of treatment time.

If continuous but precisely focused dose delivery is possible for moving tumors, total treatment time could be reduced compared with the RGRT technology in RTRT systems. Instead of RGRT, dynamic tracking radiotherapy (DTRT) technology was developed for continuous dose delivery for moving tumors with a short linear accelerator mounted on a robotic arm (CyberKnife^®)⁶⁸⁾ (Fig. 7(b)), on tomotherapy unit (Radixact^®), or on an O-ring with a gimbal (Oxray^®)⁵⁶⁾ (Fig. 7(c)), respectively. Real-time tumor tracking using external markers on the skin, internal fiducial markers, dual radiography, and prediction model, which was described in 2.3.1, was combined with the DTRT technology in these systems successfully. Akimoto et al. showed that intrafractional baseline drift requires frequent modification of the prediction model which would prolong the treatment time in DTRT.⁵⁸⁾ Pepin et al. suggested that prediction models with dynamic multileaf collimators would be useful to adjust to the intrafractional baseline drift in DTRT.⁶⁹⁾ Synchrony^® on Radixact^® uses dynamic multileaf collimators²³⁾ and has ability to correct for baseline drift in DTRT within a short treatment time in selected patients.⁵⁷⁾^,⁷⁰⁾ Yamazaki et al.⁷¹⁾ and Harada et al.⁷²⁾ pointed out that a potential drawback of the 4DRT using DTRT is the difference between the distance from the tumor to the fiducial marker in exhalation phase and the distance in inhalation phase. When the respiration cycle was divided into 10 phases, the median distance between the presumed position of the internal marker and the center of gravity of a tumor was larger for 30%–70% (inhale) of the respiratory phases compared to that for the 10% (end of exhale) of the respiratory phase (P < 0.05, Mann-Whitney U-test) (Fig. 8(a)).⁷¹⁾ Khadige et al. suggested that if the distance of the tumor from the fiducial marker is >50 mm, tumor recurrence was more common in DTRT for small lung tumors (95% vs. 69%, P = 0.011).⁷³⁾

Fig. 8.

Fiducial marking issues for the lungs. (a) The marker/tumor misalignments as a function of the respiratory phase. The horizontal lines represent the median value for the respiratory phases.⁷¹⁾ (b) Kaplan-Meier curve of fixation rate of markers from the date of insertion according to the location determination of each marker. Markers in the left upper lobe had a lower fixation rates than those in the left lower lobe (p = 0.01).⁷⁴⁾ (Reproduced from Radiat. Oncol. 7, 218 (a) and Int. J. Radiat. Oncol. Biol. Phys. 63, 1442–1447 (b).)

2.3.3. Fiducial marking.

The positional relationship between the gold markers and the tumor may change depending on the migration of the gold markers and shrinkage/deformation of the tumor, it is mandatory to check this relationship on each treatment day both in RGRT and DTRT. It was shown that migration of the marker is not rare in the lungs if bronchoscopic insertion is used (Fig. 8(b)),⁷⁴⁾^,⁷⁵⁾ although marker position is stable in the liver and prostate gland.⁷⁶⁾ The daily checking and adaptive tasks, in addition to the invasiveness, remain as disadvantages for the marker-based approach. There are several approaches to improve the stability of fiducial markers⁷⁵⁾^,⁷⁷⁾ and to reduce the invasiveness in marker implantation including injectable gold markers using nano-technology, which is one candidate for breakthroughs in this shortcoming.⁷⁸⁾^,⁷⁹⁾

Only a small proportion of lung cancers is visible on the fluoroscopic images and markerless lung target tracking with fluoroscopy has not been sufficiently robust to be used in the majority of patients up to now.⁸⁰⁾ For abdominal cancers such as the liver and pancreas, internal fiducial markers have been mandatory in fluoroscopy-based tumor tracking up to now.

2.3.4. Markerless volumetric real-time tracking.

Magnetic resonance guided radiotherapy (MRgRT) systems combining MRI equipment with radiotherapy equipment are expected to realize a markerless volumetric 4DRT (Fig. 7(d), (e)). These systems are developed and have been used in clinics from the 2010s⁸¹⁾^–⁸³⁾ after being wished for a long period.²⁹⁾ Using 4Hz 2D cine MR imaging during radiotherapy, the RGRT technology is used for immediate focus of the radiation to the moving tumors.⁸⁴⁾ The MRgRT has started to be used clinically worldwide recently.⁸⁵⁾ The daily check and adaptive task for the volumetric information is much more critical in the markerless approach than in the marker-based approach. Image reconstruction followed by automatic segmentation has not been quick and accurate enough to call it an ideal 4DRT for tumors moving with respiratory or cardiac motion.⁸⁶⁾^,⁸⁷⁾ Due to the electron return effect during MRgRT, moving gas in the digestive tract can result in over-dosage to the walls of the digestive tract.⁸⁸⁾ Reduction of system latency is a challenging issue for MRgRT, and prediction models for internal/external correlation used in fluoroscopy-based 4DRT are used to mitigate the problems associated with the latency in MRgRT.⁸⁹⁾ Real-time 3D contouring of the tumor and organs at risk (OARs) from MRI images using artificial intelligence is expected to provide breakthroughs for markerless 4DRT.³⁸⁾^,⁹⁰⁾

2.3.5. Compatibility with particle beam therapy.

Particle beam therapy such as PBT and CBT are more sensitive to uncertainties in tumor position since the path length of the particle can significantly change with organ motion.⁹¹⁾ Using RGRT technology, spot-scanning PBT with 2 sets of fluoroscopes on the gantry has been developed and named as real-time-image gated proton therapy (RGPT). It was approved by regulatory agencies in 2014 by PMDA, in 2018 by FDA, and in 2020 by MDD (PROBEAT-RT^®, Hitachi, Tokyo) (Fig. 7(f)).⁹²⁾ There had been gated PBT and CBT systems using surface surrogates or 2D imaging for gating.⁹³⁾^,⁹⁴⁾ Minohara et al. reported that, using x-ray fluoroscopy along the carbon-beam axis, they verified by sight whether the moving target and/or markers remain within the outline of the collimator of CBT during the gate-on.⁹³⁾ The RGPT is regarded as an initial 4DRT system using particle beams.⁹⁴⁾^–⁹⁸⁾ A 4D dose evaluation using trajectory data of 35 patients with a peripheral small lung tumor showed that sufficient dose homogeneity and conformity of target volume were achieved with RGPT for a gating window of ±2 mm.⁹⁹⁾ In 168 fractions in 8 liver tumors, the mean difference between the treatment planning and the actually delivered dose, which was calculated based on log data in the RGPT system, was below 1% in the metrics for dose homogeneity and conformity.⁹⁷⁾ Prolongation of beam delivery time was 2.9 ± 0.4 m (non-gated PBT) to 5.3 ± 1.3 m (RGPT) for the prostate, 2.8 ± 1.5 m to 8.7 ± 5.6 m for the liver, 3.7 ± 1.3 m to 9.7 ± 3.1 m for the lungs, 2.5 ± 0.4 m to 13.2 ± 7.0 m for the pancreas, and 1.7 ± N.A. to 5.6 ± N.A m for the adrenal tumors in 118 patients.¹⁰⁰⁾ Beam delivery efficiency was 61.8 ± 24.3% in total and the size of the intra-field adjustment corresponding to the baseline shift/drift was 3.0 ± 2.0 mm.¹⁰⁰⁾

2.4. Nonproprietary nomenclature for high-precision 3DRT and 4DRT.

The World Health Organization has established rules for international nonproprietary names for pharmaceutical substances.¹⁰¹⁾ However, international rules of nomenclature regarding radiotherapy are yet to be established by public organizations. Lacking a basis for objective discussion based on the biomedical characteristics, EBRT is often named by the mechanical classification (gantry-based, robotic, O-ring, etc.) or a commercial name in scientific papers. Imaging modalities such as MRI for the precise localization is also often used to represent the EBRT system. Although names such as IMRT, IGRT, MRgRT, PBT, and CBT are commonly used, none of the names are independent and cannot be categorized. In order to look back the biomedical outcomes of high-precision 3DRT and 4DRT objectively, tentative rules for nonproprietary naming are introduced in the following review (Table 1).

Table 1. Examples of nonproprietry name for 4DRT and their performance derived or estimated from the literature

No	Nonproprietary Name	Commercial Name	Real-time Tumor Tracking				Immediate Focused Irradiation
			Monitoring Equipment	Monitoring Interval (seconds)	Surface Surrogate	Prediction Model	System Latency (seconds)*
			Reference Points	Monitoring Interval (seconds)	Surface Surrogate	Prediction Model	Beam-on Latency	Beam-off Latency
1	0.1s-4DXT (m+0.03sDF)	Mitubishi-Linacs^® (RTRT system)	dual kV fluoroscopy	0.033⁵²⁾	−	±	0.09³¹⁾
1	0.1s-4DXT (m+0.03sDF)	Mitubishi-Linacs^® (RTRT system)	internal fiducial markers	0.033⁵²⁾	−	±	0.09³¹⁾
2	0.1s-4DXT (m+0.03sDF)	TrueBeam^® + SyncTraX^®	dual kV fluoroscopy	0.033⁵⁰⁾	−	±	0.109 ± 0.012⁵⁰⁾ (6MV-FF)	0.072 ± 0.011⁵⁰⁾ (6MV-FF)
2	0.1s-4DXT (m+0.03sDF)	TrueBeam^® + SyncTraX^®	internal fiducial markers	0.033⁵⁰⁾	−	±	0.109 ± 0.012⁵⁰⁾ (6MV-FF)	0.072 ± 0.011⁵⁰⁾ (6MV-FF)
3	0.1s-4DXT (m+30sDR+p)	CyberKnife^® (with Synchrony^®)	dual kV radiography	30⁵³⁾	+	+	0.192 (version 5.2.1)⁵⁴⁾ 0.115 (version 6.2.5)⁵⁴⁾
3	0.1s-4DXT (m+30sDR+p)	CyberKnife^® (with Synchrony^®)	internal fiducial markers	30⁵³⁾	+	+	0.192 (version 5.2.1)⁵⁴⁾ 0.115 (version 6.2.5)⁵⁴⁾
4	0.1s-4DXT (m+1sDF+p)	Oxray^® ± Anzai belt^®	dual kV fluoroscopy	1⁵⁵⁾	+	+	0.066 (Vero^®)⁵⁶⁾
4	0.1s-4DXT (m+1sDF+p)	Oxray^® ± Anzai belt^®	internal fiducial markers	1⁵⁵⁾	+	+	0.066 (Vero^®)⁵⁶⁾
5	0.1s-4DXT (m+7.5sSF+p)	Radixact^® (with Synchrony^®)	sequential kV radiography	5.3–13.9⁷⁰⁾ (median 7.5)	+	+	0.07⁵⁷⁾
5	0.1s-4DXT (m+7.5sSF+p)	Radixact^® (with Synchrony^®)	±internal fiducial markers	5.3–13.9⁷⁰⁾ (median 7.5)	+	+	0.07⁵⁷⁾
6	0.1s-4DXT (m+0.03sEMT)	Trilogy^® + Calypso^®	EMT	0.033⁶²⁾	−	−	0.075 ± 0.013⁶²⁾	0.065 ± 0.013⁶²⁾
6	0.1s-4DXT (m+0.03sEMT)	Trilogy^® + Calypso^®	internal fiducial markers	0.033⁶²⁾	−	−	0.075 ± 0.013⁶²⁾	0.065 ± 0.013⁶²⁾
7	0.1s-4DXT (0.25sMRI)	MRIdian^®	0.3 T MRI (2D-cine, 1 plane)	0.25⁸⁴⁾ 0.125⁸⁴⁾	−	±	0.342–0.464⁸⁴⁾ (0.25sMRI)	0.128–0.243⁸⁴⁾ (0.25sMRI)
7	0.1s-4DXT (0.25sMRI)	MRIdian^®	internal anatomical area	0.25⁸⁴⁾ 0.125⁸⁴⁾	−	±	0.342–0.464⁸⁴⁾ (0.25sMRI)	0.128–0.243⁸⁴⁾ (0.25sMRI)
8	0.1s-4DXT (0.25MRI)	Unity^®	1.5 T MRI (2D-cine, 3 planes)	0.2⁸¹⁾	−	±	0.125 ± 0.011⁸⁵⁾
8	0.1s-4DXT (0.25MRI)	Unity^®	internal anatomical areas	0.2⁸¹⁾	−	±	0.125 ± 0.011⁸⁵⁾
9	0.1s-4DPT (m+0.03sDF)	PROBEAT-RT^®	dual kV fluoroscopy	0.033⁹²⁾	−	±	0.13^†	0.066⁹²⁾
9	0.1s-4DPT (m+0.03sDF)	PROBEAT-RT^®	internal fiducial markers	0.033⁹²⁾	−	±	0.13^†	0.066⁹²⁾

*The definitions of system latency differ among publications. 4DXT: four-dimensional x-ray therapy, m: marker, DF: dual fluoroscopy, FF: flattening filter, DR: dual radiography, p: prediction model, SF: sequential fluoroscopy, EMT: electromagnetic transponder, MRI: magnetic resonance imaging, †: personal communication.

For the 1st requirement in high-precision 3DRT, the technology that captures the 3D coordinates of the patient tumor and its surrounding normal tissues at the start of irradiation with an accuracy of ±1 mm, examples are: cone-beam CT (CBCT), MRI, US imaging, and internal fiducial markers (m) with dual radiography (DR) have been widely used clinically. In this article, the 3DRT is named as 3DRT(CBCT), 3DRT(MRI), 3DRT(US), and 3DRT(m+DR), respectively. If the frequency of capturing 3D coordinates with CBCT is 30 times during the treatment period, the 3DRT is named 3DRT(CBCT)₃₀. If a stereotactic body frame and DR is used for the set-up, it can be expressed as 3DRT (body frame+DR). The second requirement for the 3DRT is technology for the precise focused irradiation to the static target with an accuracy of ±1 mm. If the accuracy of the 3DRT(CBCT)₃₀ is about ±1 mm for the static target, the RT is named 1mm-3DRT(CBCT)₃₀.

Regarding the 1st requirement for 4DRT, the minimum frequency of monitoring can differ with the speed of tumor motion to term it “real-time”. Approximately, “real-time” can be used for two types of tumor motion: a slow-type motion which requires monitoring approximately every 1 s or less, and fast-type motion which requires monitoring every 0.1 s, respectively. Bowel motion and global body motion during radiotherapy are slow-type motions and respiratory and cardiac motion are fast-type motions, generally speaking.¹⁰²⁾ Internal fiducial markers with dual X-ray fluoroscopies (DF), sequential fluoroscopy (SF), implantation of EMT with the radiolucent antenna array positioned just above the patient,¹⁰³⁾ and MRI have been used for both types of tumor motion in the clinic. Metallic markers with DF at every 0.03 s, EMT every 0.03 s, and MRI every 0.25 s are termed 4DRT(m+0.03sDF), 4DRT(0.03sEMT), and 4DRT(0.25sMRI), respectively. Monitoring of surface surrogates on chest or abdominal wall with infrared light with a prediction model (p) based on the internal fiducial marker position certified with DR every 30 s, and with SF every 7.5 s are termed 4DRT(m+30sDR+p) and 4DRT(m+7.5sSF+p), respectively.

Regarding the 2nd requirement for 4DRT, technology for immediate focused irradiation to the moving tumor, the meaning of “immediate” can also differ with the speed of tumor motion. Generally speaking, there are two types of acceptable latencies for immediate irradiation, slow-type (approximately within 1 s) and fast-type (approximately within 0.1 s) respectively. The slow-type latency was common in photon EBRT with diode electron guns and in conventional PBT and CBT to maintain dosimetric reliability but the fast-type latency became common in recent photon EBRT and scanning PBT and CBT. A 4DRT, in which the beam is gated with the latency of about 0.1 s, is termed 0.1s-4DRT.

The 3DRT is called 3DXT or 3DPT if the therapeutic beam is an X-ray or proton beam respectively in this review. The same role applies to 4DRT. If any one of the requirements for high-precision 3DRT and 4DRT are not fulfilled, the RT has a lower dimensionality irrespective of the mechanical preciseness of the medical devices. However, there are adaptive approaches for very slow, daily, or weekly, anatomical changes during EBRT due to shrinkage of the tumor mass, weight loss of the patient, or for other reasons. Recent high-precision 1mm-3DRT(MRI) can use the 3D coordinates of the organs and tumors at each treatment day for dose calculations and the actual treatment beam is focused to tumors which change position and shape slowly. These “anatomical slow change” situations are different from the classic “organ motion” but these EBRT are sometimes referred as 4DRT broadly considered, recently.¹⁰⁴⁾ For example, online adaptive EBRT with 30 minutes required for adaptation to the daily position or shape may be termed 30m-4DRT. If they use an isotoxic planning policy (IPP) in which a re-treatment plan is executed when there are errors of more than a threshold on the daily MRI, it may be called 30m-4DRT(MRI)/IPP.¹⁰⁴⁾ Table 1 shows examples of 4DRT systems using the rules for nonproprietary naming. The performance in each column is subject to the limitation that they may have changed after the publication of original papers.

3. Biological advances in high-precision 3DRT and 4DRT

3.1. Normal tissue complications.

It is obvious that high-precision 3DRT and 4DRT reduced the volume of the normal tissue which cannot be shielded from high doses and thus reduces the risk of radiation-induced adverse effects to lower levels than with 2DRT. However, there is still the question of whether the radiation dose, absorbed energy in a unit of weight (J/Kg = Gy), to cancer could have increased by high-precision 3DRT and 4DRT. If the dose to the cancer can be increased by high-precision 3DRT and 4DRT, we can expect improvement in TCP and thus a better survival rate of patients. However, since most of the cancer has already invaded into the surrounding normal tissue when cancer is diagnosed, we cannot increase the dose to the cancer without increasing the dose to the surrounding normal tissue. Therefore, the real question is whether we were able to increase the dose to the normal tissue surrounding the cancer without serious complications when we decrease the irradiated volume of the normal tissue by the high-precision 3DRT and 4DRT. So, “whether we could increase the absorbed dose in a part of the OAR maintaining the NTCP of the OAR below the threshold for causing adverse effect by reducing the irradiated volume to the normal tissue” is the real question need to be answered before introducing high-precision treatment. It is often rephrased as whether there is a “volume effect” in the normal tissue.¹⁰⁵⁾

To predict the answer to this question, mathematical models for NTCP assuming that an OAR is composed of numerous independent critical segments (CSs) was proposed in late 1980s when the shift from 2DRT to 3DRT occurred.¹⁰⁵⁾^,¹⁰⁶⁾ Each OAR is categorized into parallel- or serial-types in respect to the structure of the CS (Fig. 9(a), (b)).¹⁰⁷⁾ Typical parallel-type OAR are the parenchyma of the lung, liver, and kidneys. Typical serial-type OAR are the spinal cord and digestive tract. If we accept these assumptions, a radiation-induced adverse end-point will occur when a certain number of CS are injured in parallel-type OAR. Differently, a radiation-induced adverse end-point will occur when any one of the CS are injured after radiotherapy in serial-type OAR. There are various models to calculate the NTCP statistically assuming dose-dependent, threshold-binary, stochastic events in each CS after radiotherapy.¹⁰⁷⁾ Purely mathematical or phenomenological models to fit a sigmoid shape of NTCP have also been used widely.¹⁰⁸⁾^–¹¹¹⁾ Since the dose distribution in OAR is usually inhomogeneous in 3DRT, the prediction of NTCP became practical in the 1990s when the absorbed dose in each CS could be computed precisely. The dosimetric indices theoretically suitable for the prediction of NTCP and TCP in the 3DRT era, which are used as the unit on the horizontal axis in Fig. 1(f), were found to be substantially different for the tumor and OAR.¹¹¹⁾ Since any residual CS after radiotherapy can be a source of relapse of cancer, the tumor can be regarded as an “organ” with the serial-type structure. The minimum dose (D_min) was found to be the metric to predict TCP since one surviving CS can be the source of relapse. When we can increase the tumor D_min keeping the dosimetric indices of the OAR below the threshold by high-precision 3DRT or 4DRT, we can increase cure rates without increasing adverse effects even when each one of the CS in the OAR is more radiosensitive than the CS in the tumor (Fig. 1(f)).

Fig. 9.

Illustration of (a) a parallel-type organ and (b) a serial-type organ: Each critical segment (CS) is shown as blue box and the radiation field is shown as orange box. The radiation field contains n CS.¹⁰⁵⁾^–¹⁰⁷⁾ (c) Representation of the cell migration model.¹³³⁾

In the following sections, we review biomedical advances using published animal experimental results and human clinical outcomes. With the improvements in the preciseness of tumor positioning in EBRT, fewer fractionations of absorbed doses in shorter periods have become common to reduce burden on both patients and medical staff with the hope to improve cure rates for some types of cancer.²¹⁾^,¹¹²⁾^,¹¹³⁾ To predict clinical outcomes of EBRT with fewer fractionations, biological evaluation metrics, such as the biological effective dose,¹¹²⁾ or the equieffective dose (EQD),¹¹⁴⁾ have been developed and are widely adopted clinically. In this review, we use EQD2_α/β which is assumed to be biologically equivalent to the total dose using 2 Gy as the daily dose administered five times a week. When we review the tolerance dose of OAR, EQD2₂ will be described together with the physical dose. When we review the tumor control and survival, both EQD2₂ and EQD₁₀ will be described together with the physical dose.

3.1.1. Biological advances for parallel-type OARs.

According to ordinal NTCP models, both the mean dose (D_mean) and relative volume of the OAR which receives a specific dose (d) or more (V_d), are theoretically suitable to predict NTCP in parallel-type OAR.¹¹⁵⁾^–¹¹⁸⁾ Expressed differently, the maximum dose (D_max) to a part of a parallel-type OAR can dramatically increase safely by decreasing the irradiated volume providing that the D_mean or V_d is maintained below a threshold for the parallel organ. It was confirmed that the D_mean of the lung parenchyma, a parallel OAR, is a useful index to predict the NTCP of radiation-induced pneumonitis using various patient data from the 3DRT era.¹¹⁸⁾^,¹¹⁹⁾ The D_mean of the liver parenchyma, a parallel OAR, was also shown to be a good index for radiation-induced liver damage.¹²⁰⁾^,¹²¹⁾ Therefore, high-precision 3DRT and 4DRT have been noted as possibly improving the survival rate of patients with small lung and liver cancers by increasing the absorbed dose to the small volume around the tumor maintaining the D_mean or V_d of the lung and liver parenchyma, parallel-type OAR. These show that there is a volume effect for parallel-type OAR.

3.1.2. Biological advances for serial-type OARs.

For serial-type OAR however, the D_max to a part of an OAR cannot be increased safely by decreasing the irradiated volume according to ordinary NTCP models.¹⁰⁷⁾^,¹²²⁾ This is consistent with the reported tolerable D_max of the esophagus, spinal cord, and other serial-type OARs did not change with the length of treatment field in the 2DRT era.¹²³⁾ The tolerable doses for serial-type OARs have not changed in the range of ordinary treatment lengths longer than several cm even in the 3DRT era.¹²⁴⁾^,¹²⁵⁾ Therefore, it has been believed that there is no volume effect for serial-type OAR in general. However, there have been few data points in the dose-response relationship for small volume of serial-type OAR also from the 3DRT era. Since serious adverse reactions of deep-seated serial-type OAR are not always detectable even with CT or MRI examinations, we may be overlooking some cases or over estimating adverse reactions if we use only published clinical data. In fact, there are several experimental studies in small animals and also human observations suggesting that a higher D_max can be given when the irradiation volume is reduced from approximately 20 mm to 1 mm in radius in serial-type OAR. When the irradiated length of the spinal cord of rats have been changed from 16 mm to 4 mm,¹²⁶⁾ from 40 mm to 5 mm,¹²²⁾ or from 20 mm to 2 mm,¹²⁷⁾ the dose that produces the same adverse end-point, paralysis, increased dramatically both in photon¹²²⁾^,¹²⁶⁾ and proton irradiation.¹²⁷⁾ Experiments in large animals showed increases in the tolerable D_max when the length of the irradiated peripheral nerve is reduced from 15 to 5 mm.¹²⁸⁾ In humans, stereotactic radiosurgery (SRS) for intracranial lesions, where precise irradiation compatible with the above experiment is possible, showed that the risks of symptomatic radionecrosis of the brain for volumes receiving 12 Gy, V_12Gy, in a single fraction (EQD2₂ = 42 Gy), was approximately 20% for >15 cc (15.0 mm in radius) and 10% for 5 cc (10.6 mm in radius) respectively.¹²⁹⁾ This is consistent with the clinical guidelines for dose constraints of the brain SRS, V_12Gy ≤ 5 cc.¹²⁴⁾^,¹²⁵⁾ Regarding the digestive tract, a serial-type OAR, 15 Gy using daily doses of 1.8–2 Gy (EQD2₂ = 14.3–15 Gy) for the 120 cc was the dose constraint in conventional 2DRT and 3DRT.¹²³⁾^,¹²⁴⁾ The dose constraints used in a recent prospective study for 4DRT were 35 Gy/5 Fr (EQD2₂ = 78.8 Gy) for 0.5 cc and 50 Gy/5 Fr (EQD2₂ = 150 Gy) for 0.03 cc of the stomach, duodenum, and small bowel.¹³⁰⁾ The mean of the maximum doses to 1 cc of duodenum was significantly higher in patients developing grade ≥2 ulceration or stenosis 37.4 Gy/3 Fr (EQD2₂ = 135.3 Gy) than others 25.3 Gy/3 Fr (EQD2₂ = 66 Gy) (p = 0.03).¹³¹⁾ Skin late radionecrosis was also rare at the heavily irradiated (58 Gy/1 Fr) area if the radius was 5 mm.¹³²⁾ These clinical reports and guidelines do not exclude the possibility of the existence of a volume effect from approximately 20 mm to 1 mm in radius in serial-type OAR in humans.

The possibility of the presence of a volume effect at the small volume levels in serial-type OAR is contrary to the ordinary NTCP models. However, one of the NTCP models showed that the volume effect seen in the serial-type OAR can be explained if we assume a molecular process in radiation-induced injuries which depend on dose and volume as will be explained next. It was proposed in 1995 assuming recruitment of parenchymal cells from outside of the CS (Fig. 9(c)).¹³³⁾ In this model, repopulation of the parenchymal cells in CS due to cell migration from outside of the CS increases when the dose in the irradiated field is increased, but it decreases when the volume of irradiated field increased (Fig. 9(c)). Philippens et al. have shown that the cell-migration model fitted far better than any other ordinal NTCP models although they are not certain about the appropriateness of a biological hypothesis in the cell-migration model (Fig. 10).¹²²⁾ Bijl et al. have shown that simple migration of glial progenitor cells from the edge of the treatment volume is likely not the reason for the repopulation of cells in the irradiated field.¹³⁴⁾ Recent radiation biology suggests that the recovery process is dependent on complex cell to extra-cellular matrix interactions, upregulation of cytokines and growth factors, damage to the endothelium, and recruitment of cells far from the treatment field.¹³⁵⁾^–¹³⁹⁾ These new findings are consistent with the biological hypothesis in the cell-migration model although more work is required to confirm the appropriateness of the model.

Fig. 10.

Fitted dose-response data of the 4.0-cm (+), 2.5-cm (■), 1.0-cm (◆), and 0.5-cm (▲) field lengths of a rat thoraco-lumbar spinal cord irradiation plotted together with the experimental data. The lines represent the fitted curves for the experimental data based on different NTCP models: (a) the Lyman model, (b) the critical model, and (c) the Shirato (cell migration) model. The bottom table shows the number of free parameters, the deviance, and the goodness of fit (Monte Carlo).¹²²⁾ (Int. J. Radiat. Oncol. Biol. Phys. 60, 578–590, modified with permission)

3.1.3. Biological advances for complex regions.

Most extra-cranial cancers are located at the complex regions consisting of serial- and parallel-type OAR. Even if the cancer occurred in a parallel OAR, the tumor mass is often situated near serial OAR. Therefore, the dose to cancers can be increased only if the serial-type OAR fulfil their dose constraints in addition to the parallel OAR satisfying their dose constrains. For cancers at complex regions, NTCP for at least one of the OAR will be subject to an adverse reaction (NTCP_w) is calculated using the following formula in which NTCP_j is the NTCP for j-th OAR.¹¹¹⁾ M is the number of OARs.

\begin{equation} \mathrm{NTCP}_{\text{w}} = 1 - \prod\nolimits_{\text{j}}^{\text{M}} (1 - \mathrm{NTCP}_{\text{j}}). \end{equation}

[1]

3.2. Tumor control.

The metrics such as D_min or D_99% of the tumor, which is the dose covering 99% of the tumor and represent the minimal dose to the tumor, have become important metrics in high-precision 3DRT and 4DRT.¹⁴⁾^,¹¹¹⁾ Tumor motion has been regarded as the major risk to lower the D_min or D_99% in high-precision 3DRT. Figure 2(a) shows that D_min can change drastically for tumors which move with organ motion in high-precision 3DRT. The mean of calculated D_min for 3 patients with peripheral lung cancer was 54.7 Gy, 33.5 Gy, and 6.1 Gy at the middle, at the exhalation, and the inhalation phase, respectively if we use CT during normal breathing for the planning and high-precision 3DRT with 10 mm margins in the treatment (Fig. 2(a)–(c)).¹⁴⁾ Here, 4DRT has been expected to keep D_min of the tumor high enough without increases in NTCP as shown in Fig. 2(d)–(f).

Complexity of tumor motion and its potential effect on tumor control were revealed in the clinical usage of 4DRT systems. Also pitfalls in the 4DRT concept due to the complexity of internal tumor motion have been established along with a precise analysis of patient data.¹⁰²⁾^,¹⁴⁰⁾ For example, hysteresis, the difference between the inhale and exhale trajectory of the tumor (Fig. 11(a), (b)), and a motion component due to cardiac motion in addition to the respiratory motion (Fig. 11(c)) were identified in patients with small peripheral lung tumors.¹⁴¹⁾ The amplitude of lung tumors changed depending on the location of the cancer (Fig. 11(d))¹⁴¹⁾ and the 3D movement was shown to be different for different treatment days in the same patient (Fig. 11(e)).⁵²⁾ The position of the tumor is usually stable at the end of the exhale but variable at the end of the inhale in normal breathing (Fig. 11(f)). Baseline shift/drift of the cancer position during radiotherapy is common: At the end of expiration, the baseline shift/drift exceeded 3 mm in 6.0%, 14.0%, 15.5%, and 42.1% for the LR, CC, AP and for the square-root of the sum of the 3 directions, respectively, within 10 minutes of the start of treatment, and 23.0%, 32.5%, 37.6%, and 71.6% within 30 minutes, in 68 patients with peripheral lung tumors (Fig. 11(g)).¹⁴²⁾ The amplitude of tumor motion in 20 patients with hepatic tumors was 1–12, 2–12, and 2–19 mm in the LR, CC, and AP directions respectively.¹⁴³⁾

Fig. 11.

Analysis of movement of a fiducial marker implanted near a lung tumor at a sample rate of 30 images per second. (a) The 3D path during one treatment portal. Blue dots are the position during beam-off, and the red dots are the position during the beam-on periods in 4DRT.¹⁴¹⁾ (b) The average 3D trajectory of the marker showing hysteresis: The projections on the coronal, sagittal, and axial planes are drawn in thin blue lines.¹⁴¹⁾ (c) The frequency spectrum of a marker motion during one treatment. Two frequency peaks are present, one at 0.3 Hz is a result of breathing and one at 1.05 Hz is the result of the heartbeat.¹⁴¹⁾ (d) Orthogonal projections of the trajectories of 21 tumors on (left) the coronal and (right) the sagittal plane. Tumors that were attached to bone structures are marked in orange. The lower lobe tumors are colored light blue and the upper lobe tumors purple.¹⁴¹⁾ (e) Changes of the amplitudes of the lung tumor in the craniocaudal direction in 60 treatments of 21 patients.⁵²⁾ (f) Lung tumor position at the end of exhale (blue) and at the end of inhale (red).⁵²⁾ (g) Incidence of baseline shift/drift exceeding 3 mm for the left-right (LR), cranio-caudal (CC), and antero-posterior (AP) directions, and for the square-root of the sum of the 3 directions (3D).¹⁴²⁾ (Reproduced from Int. J. Radiat. Oncol. Biol. Phys. 53, 822–834 ((b)–(d)), 64, 1229–1236 ((e), (f)), and 94, 172–180 (g).)

Respiratory motion also affects the motion of abdominal and pelvic organs. The amplitude of the prostate motion was significantly less (0.1–2.7 mm) in the supine position than in the prone position (0.4–24 mm) in the same 10 patients with prostate cancers (p < 0.0001) suggesting the respiratory motion of the prostate gland in the prone position (Fig. 12(a)).¹⁴⁴⁾ As for slow-type motion, prostate cancers also move due to rectal and bladder motion. The incidence of table position adjustment required to keep intrafractional uncertainties within 2.0 mm in either of the LR, CC, or AP directions during IMRT in the supine position was investigated with the RTRT system in 20 patients with prostate cancer.¹⁴⁵⁾ The incidence of table position adjustment at 10 minutes from the initial setup for each treatment was 5.0%, 12.3%, and 14.2%, of the 4541 observations in the LR, CC, and AP directions, respectively. Adjustments of more than 5 mm was required at least once in 10 minutes in 7 (35%) patients and at some points in the treatment period in 11 (55%) of the 20 patients entered in their study (Fig. 12(b)).¹⁴⁵⁾ Precise knowledge about the motion of a fiducial marker in or near (in/near) the tumor in other organs such as the upper digestive tract,¹⁴⁶⁾ pancreas,¹⁴⁷⁾ adrenal grand,¹⁴⁸⁾ uterus,¹⁰²⁾ and other organs¹⁴⁹⁾ were also reported using the RTRT data. Uchinami et al. recently investigated changes in contours of the digestive tract around the pancreas using diagnostic CT scans taken 3 times within a mean duration of 12 minutes (interquartile range, 10–16 minutes) (Fig. 12(c)).¹⁵⁰⁾ They found that a median margin of 8.0–14.0 mm was necessary to compensate even for the short-term uncertainties of the position of the abdominal digestive tracts. Their results suggest that even though the mean doses to the digestive tract do not change very much, the D_max of each digestive tract, which is the metrics suitable for serial-type OAR, is ever-changing and difficult to calculate accurately in real-time even with the fast MRI at present.

Fig. 12.

(a) 3D trajectory of a fiducial marker at the apex of the prostate gland for 2 min in 4 patients (left) supine position and (right) prone position.¹⁴⁴⁾ (b) Cumulative incidence and the magnitude of displacement of patient-table position adjustments (vertical, longitudinal, and lateral) during treatments in 20 patients with prostate cancer showing increases in displacement with time.¹⁴⁵⁾ The incidence in each category is shown as the matrix of 2 mm<, 4 mm<, 6 mm<, 8 mm<, and 10 mm< adjustment and <2 min, <4 min, <6 min, <8 min, <10 min, and 10 min< from the start of radiotherapy. (c) The contours of the stomach (blue), duodenum (pink), small intestine (green), and large intestine (purple) at 3 times during 12 minutes simultaneously superimposed on a CT image, which indicates variations of the contours in 12 minutes.¹⁵⁰⁾ (Reproduced from Int. J. Radiat. Oncol. Biol. Phys. 53, 1117–1123 (a), 81, e393–e399 (b), and Clin. Transl. Radiat. Oncol. 39, 100576 (c).)

4. Medical advances in high-precision 3DRT and 4DRT

4.1. Cancers in parallel-type OAR.

From the late 1990s, high-precision 3DXT has been used to treat small cancers in parallel OAR. Very high, ablative doses have been used for a small cancer visible on a CT including an about 10 mm margin around the tumor.¹⁵¹⁾^–¹⁵³⁾ The high-precision 3DXT using small-volume giving high, ablative doses for extra-cranial lesions has been termed stereotactic body radiation therapy (SBRT).¹⁵¹⁾ Arimoto et al. used orthogonal portal images in the treatment room and reported that SBRT, or 1mm-3DXT(DR)₈, given 60 Gy in 8 fractions (EQD2₂ = 142.5 Gy, EQD2₁₀ = 87.5 Gy) in 11 days for the 24 lung tumors < 15 mm in radius and reported a 91.7% 2-year local control rate (LC) with 5% mild adverse events (cough lasts for a few weeks) in 1998 (Table 2).¹⁵²⁾ Uematsu et al. verified the results of Arimoto for stage I non-small cell lung cancer (NSCLC) using on-line CT in the treatment room for the daily set-up of the patients,¹⁵³⁾ or 1mm-3DXT(online CT)_5–10, giving 50–60 Gy in 5–10 fractions (50–60 Gy/5–10 Fr) over 1–3 weeks.¹⁵⁴⁾ The 3-year overall survival rate (OS) was 66% with mild toxicity in the Uematsu study (Table 2).¹⁵⁴⁾ Since the 3-year OS of patients with medically inoperable stage I NSCLC in conventional 2DRT was 17–55% in the early 1990s,¹⁵⁵⁾ these results with SBRT were encouraging. Results of 4DXT either 0.1s-4DXT (m+0.03sDF) or 0.1s-4DXT (m+30sDR+p) giving ablative doses for peripheral stage I NSCLC were similar to the high-precision 3DXT (Table 2).¹⁵⁶⁾^,¹⁵⁷⁾ The 5-year OS was 64% [95% confidence interval (CI) 53–78%] in 0.1s-4DXT (m+0.03sDF).¹⁵⁶⁾ A recent phase I study of 1mm-3DXT (miscellaneous) or ≤1s-4DXT (miscellaneous) for peripheral stage I NSCLC showed that the tolerable D_max of the lung parenchyma did not have a ceiling in situations where the lung V_20Gy was lower than the threshold.¹⁵⁸⁾ A multi-institutional retrospective study showed that ablative doses (EQD2₂ ≥ 157.5 Gy, EQD2₁₀ ≥ 83.4 Gy) given with 1mm-3DXT (miscellaneous) or ≤1s-4DXT (miscellaneous) was associated with response rates (RR) of 84.5% and 5-year OS of 47%, which were higher than reported results of conventional 2DXT (RR 40–70% and 5-year OS 6–32%) for peripheral stage I NSCLC.¹⁵⁵⁾^,¹⁵⁹⁾ A recent RCT for peripheral stage I NSCLC confirmed that 1mm-3DXT (CBCT or m+DR) giving ablative dose (54 Gy/3 Fr or 48 Gy/4 Fr) (EQD2₂ = 168 or 270 Gy, EQD2₁₀ = 88 or 126 Gy) resulted in superior LC (2-year LC 89%) compared with conventional 2DXT or 3DXT (66 Gy/33 Fr or 50 Gy/20 Fr) (EQD2₂ = 66 or 56.3 Gy, EQD2₁₀ = 66 or 52.1 Gy) (2-year LC 65%) (HR of 0.32 [95% CI 0.13–0.77], p = 0.008) without an increase in toxicity (Table 2) (Fig. 13(a)).¹⁶⁰⁾ Median survival time in the 1mm-3DXT(CBCT or m+DR) group was 5 years which was significantly longer than the 3 years in the conventional 2DXT or 3DXT groups (HR 0.53 [95% CI 0.30–0.94], p = 0.027) in the RCT.¹⁶⁰⁾ Clinical guidelines at present recommend high-precision 3DXT or 4DXT for patients with peripheral Stage I NSCLC who are medically inoperable or when patients decline to receive surgery.¹⁶¹⁾ Furthermore, for operable peripheral stage IA NSCLC, Onishi et al. retrospectively showed that 1mm-3DXT (miscellaneous) or ≤1s-4DXT (miscellaneous) have achieved a 5-year OS of 72%, which was comparable to surgery (Table 2) (Fig. 13(b)).¹⁶²⁾ Following the Onishi et al. report, a prospective phase II study of 1mm-3DXT (miscellaneous) or ≤1s-4DXT (miscellaneous) was conducted for peripheral stage IA NSCLC achieving a 3-year OS of 76.5% [95% CI 64.0%–85.1%] for operable patients (Table 2).¹⁶³⁾ Recently, pooled estimated OS at 3 years of 1mm-3DXT (CBCT or in-room CT) for operable stage I NSCLC was reported to be as high as 95% [95% CI 85–100%] favorably comparing with the 79% [95% CI 64–97%] in the surgery group (HR of 0.14 [95% CI 0.017–1.190], p = 0.037) (Fig. 13(c)).¹⁶⁴⁾ However, it is not conclusively shown whether high-precision 3DXT or 4DXT is equivalent to or better than surgery for operable peripheral stage I NSCLC, mainly because of the difficulty to do RCT. It should also be noted that there has been no RCT-based evidence which showed 4DXT as better than high-precision 3DXT for peripheral stage I NSCLC, despite the theoretical advantage of 4DXT.

Table 2. High-precision 3DRT and 4DRT for Stage I non-small cell lung cancer

Nonproprietary Name	Names in the Literature	PTV Margin for CTV (mm)	Dose Fractionation (Gy/Fractions)	Adverse Events (Grade, G)	Local Control (years, yrs)	Overall Survival (OS)	Remarks
Inoperable and Operable Stage I NSCLC
1mm-3DXT (DR)₈	SMART	not available	60 Gy/8 Fr	G 2 5%	Crude 2-yrs 91.7.%	3-yrs OS 60.4%	Arimoto et al. 1998¹⁵²⁾ 40 patients including 15 metastasis
1mm-3DXT (online CT)_5–10	FOCAL	5–10 mm	50–60 Gy/ 5–10 Fr (+ 40–60 Gy/20– 33 Fr in 18 pts)	Bone fracture 4% G 1 pleural pain 12%	94% Median follow-up 36 months	3-yrs OS Total 66% Operable 86%	Uematsu et al. 1998¹⁵³⁾ 50 patients including 29 operable patients
0.1s-4DXT (m+0.03sDF)₄	RTRT	5 mm	48 Gy/4 Fr	G 3 2.8%	3-yrs 81% 5-yrs 78%	3-yrs OS 68% 5-yrs OS 64%	Inoue et al. 2013¹⁵⁶⁾ 109 patients
0.1s-4DXT (m+30sDR+p)₃	CyberKnife^® with Synchrony^®	5 mm to GTV	45 Gy/3 Fr 60 Gy/3 Fr	G 3 10%	2-yrs 78% for 45 Gy 2-yrs 96% for 60 Gy	2-yrs OS 62%	Van der Voort van Zyp et al. 2009¹⁵⁷⁾ 70 patients
1mm-3DXT (CBCT or m+DR)_3–4	miscellaneous	5 mm LR/ AP, 10 mm CC for ITV	54 Gy/3 Fr, 48 Gy/4 Fr	G 3 10.6% G 4 1.5%	2-yrs 89%	2-yrs OS 77%	Ball et al. 2019¹⁶⁰⁾ 66 patients
Operable Stage IA and IB NSCLC
1-mm-3DXT (miscellaneous) ≤1s-4DXT (miscellaneous)	miscellaneous	7–15 mm	45–72.5 Gy/ 3–10 Fr	G 3–4 Pneumonitis 2.4% Esophagus 0.8% Dermatitis 0.8%	Response rate 81.7% CR 32.3% PR 49.4%	5-yrs OS Stage IA 72% Stage IB 62%	Onishi et al. 2011¹⁶²⁾ 87 patients
1mm-3DXT (miscellaneous) ≤1s-4DXT (miscellaneous)	miscellaneous	5 mm for ITV	48 Gy/4 Fr	G 3 6.2%	3-yrs 85.4% [95% CI: 73.8– 92.1%]	Stage IA 3-yrs OS 76.5% [95% CI: 64– 85%]	Nagata et al. 2015¹⁶³⁾ 64 patients
1mm-3DXT (CBCT or in-room CT)	Varian 2100^®, TrueBeam^®	5 mm for ITV	54 Gy/3 Fr, 50 Gy/4 Fr	G 3 1%	Stage IA 5-yrs 93.7% [95% CI: 86.8– 97.7%]	Stage IA 3-yrs OS 91% [95% CI: 85– 98%]	Chang et al. 2021¹⁶⁴⁾ 80 patients

PTV: planning target volume, CTV: clinical target volume, NSCLC: non-small cell lung cancer, 3DXT: three-dimensional x-ray therapy, 4DXT: four-dimensional x-ray therapy, DR: dual radiography, CBCT: cone-beam computed tomography, CT: computed tomography, DF: dual fluoroscopy, m: marker, p: prediction model, GTV: gross tumor volume, LR: left-right, AP: antero-posterior, CC: cranio-caudal, ITV: internal target volume, CR: complete response, PR: partial response.

Fig. 13.

Overall survival curves of patients with stage I non-small cell lung carcinomas (a) Patients were treated with RT either using conventional 2DRT or 3DRT (blue) or 1mm-3DRT (CBCT or m+DR) (red) in a phase 3, open-label, randomized controlled trial. Modified with Permission¹⁶⁰⁾ (b) Operable patients treated with 1mm-3DRT (miscellaneous) or ≤1s-4DRT (miscellaneous) plotted by cancer stage:1A (≤3 cm) and 1B (>3 cm).¹⁶²⁾ (c) Operable patients treated with 1mm-3DRT (CBCT or in-room CT) or surgery in a single-arm prospective trial with prespecified comparisons to surgery.¹⁶⁴⁾ (Reproduced from Lancet Oncol. 20, 494–503 (a), 22, 1448–1457 (c), and Int. J. Radiat. Oncol. Biol. Phys. 81, 1352–1358 (b) modified with permission.)

In the 2DRT era, only palliative doses could be given to liver cancer because it was not possible to identify the position of the cancer precisely and hepatic normal cells are usually more radiosensitive than cancer cells. With 1mm-3DXT (customized vacuum-pillow and offline CT) using ablative doses, 96.3% LC and 66.7%OS at 3 years was achieved for 90 patients with solitary peripheral hepatocellular carcinomas (HCC) 20 mm or smaller in radius without life-threatening adverse events.¹⁶⁵⁾ The 5-year OS was 60.4% [95% CI 47.0%–73.8%] for 81 patients with HCC, who received 1mm-3DXT (miscellaneous) 2 times or more with the median prescribed dose 40 Gy/5 Fr (EQD2₂ = 100 Gy, EQD2₁₀ = 60 Gy), in a multi-institutional retrospective study.¹⁶⁶⁾ The frequency of grade (Gr) 3 toxicity was 11% [95% CI 5.2%–20%] at the first radiotherapy and 15% [95% CI 7.9%–24%] at the second.¹⁶⁶⁾ Earlier results of 0.1s-4DXT (m+0.03sDF) showed similar LC in which the 2-year LC of 15 patients with a mean diameter of 36 mm was 83% with Gr 3 gastric ulcers in one patient and a Gr 3 transient increase of aspartate amino transaminase in 2 patients.¹⁶⁷⁾ A prospective study of 0.1s-4DXT (m+1sDF+p)₄ giving 40 Gy/4 Fr (EQD2₂ = 120 Gy, EQD2₁₀ = 66.7 Gy) also showed similar efficiency and toxicity (2-year LC of 98.0% with Gr 3 late adverse events in 14.5%) in 48 patients with liver cancers with the mean diameter of 17.5 mm.¹⁶⁸⁾ In summary, adverse events due to radiation-induced liver damage were rare for small liver cancers with either of high-precision 3DXT or 4DXT.

Irrespective of treatment with high-precision 3DXT or 4DXT, photon therapy using X-rays cannot give curative doses to liver cancers of approximately 60 mm or more in diameter since the mean liver dose would exceed the threshold of liver damage.¹⁶⁹⁾ However, high-precision 3DRT or 4DRT using PBT and CBT can give sufficient doses to solitary cancers of 60 mm or more in diameter without the risk of increasing the D_mean of the liver, thus, the risk of radiation-induced liver damage theoretically using the NTCP model.¹⁶⁹⁾ Sugahara et al. have reported that LC for HCC with a median diameter of 110 mm (range, 100–140 mm) was 87% at 2-years with no late Gr 3 adverse events after PBT.¹⁷⁰⁾ Recent a multi-institutional prospective registration study of PBT and CBT with high-precision 3DRT or 4DRT technology showed 34.6% [95% CI 25.9–43.5%] for HCC 50–100 mm and 30.6% [95% CI 19.6–42.3%] 3-years OS for HCC 100 mm or larger in diameter.¹⁷¹⁾

Theoretically, 4DRT can reduce the D_mean of the parallel-type OAR significantly comparing to 3DRT, but there has been no proper RCT which examined the superiority of 4DRT over high-precision 3DRT for cancers in parallel-type OAR. Chapman et al. have compared, in the computer simulation, the D_mean of lung parenchymas with 4DXT using smaller margins and 3DXT using wider margins for the same 20 lung cancers with the mean diameter of 35 mm (21 to 55 mm) (Fig. 14(a)).¹⁷²⁾ The mean lung dose was significantly lower for 4DXT with a median decrease in 1.9 Gy (range 0.22–7.37 Gy) (p < 0.001). They also showed the superiority of 4DXT in various V_d of lungs (Fig. 14(b)). However, the amount of decrease in NTCP was negligible (<2%) for radiation-induced pneumonitis for half of the patients and a >5% increase of NTCP occurred only for 25% of the patients.¹⁷²⁾ It was concluded that the risk reduction in the NTCP for radiation-induced pneumonitis by photon 4DXT is minimal for most patients with peripheral small lung cancers compared to high-precision photon 3DXT. Kanehira et al. have shown that there is a similar relationship in PBT for lung V_20Gy.⁹⁹⁾ The RGPT with an appropriate gating window (1–2 mm) achieved significant reductions of lung V_20Gy but the amount of improvement compared with non-gated PBT was small (Fig. 14(c)). However, it was noted that RGPT improved the minimal dose in the tumor (D_99%) compared with non-gated PBT markedly (Fig. 14(d)),⁹⁹⁾ suggesting that higher TCP can be achieved with RGPT.

Fig. 14.

(a) Comparison of % dose profile on a coronal slice of CT using (upper) 4DXT and (lower) 3DXT technology for one patient with lung tumor to expose the tumor to a sufficient dose. Modified with permission from Ref. 172. (b) Boxplots of lung V_2.5Gy, V_5Gy, V_10Gy, V_13Gy, V_20Gy, V_30Gy, V_40Gy, and V_50Gy (%) by (red) high-precision 3DXT and (green) 4DXT.¹⁷²⁾ (c) Box plots of V_20Gy (%) of the normal lung, (d) Box plots of D_99% of the tumor for 7 patients with peripheral small lung cancer comparing 4D PBT with gating window [GW] sizes, and 3D PBT in free-breathing [FBPT]. Boxes show the interquartile range (IQR) from the first (Q1) to the third quartile (Q3). Stars indicate outliers (data below Q1 − 1.5 × IQR or above Q3 + 1.5 × IQR).⁹⁹⁾ (Reproduced from J. Appl. Clin. Med. Phys. 19, 48–57 ((a), (b)) and Int. J. Radiat. Oncol. Biol. Phys. 97, 173–181 ((c), (d)).)

In summary, high-precision 3DRT is medically superior to 2DRT for parallel-type OAR in general. Since there is a volume effect in the parallel-type OAR, if the D_mean of the parallel-type OAR does not change greatly, we can increase the maximal dose to a part of the parallel-type OAR safely with high-precision 3DRT. Differences in lung D_mean are small between high-precision 3DRT and 4DRT for the majority of small peripheral cancers in parallel-type OAR. We can increase the TCP of peripheral small cancers either with high-precision 3DXT or 4DXT in the parallel-type OAR. As a medical improvement, we have better results for the LC and probably the OS of patients who had peripheral small cancers in parallel-type OAR by either high-precision 3DXT or 4DXT in the past 20 years. Since tumor D_min can be improved by a combination of 4DRT technology with PBT and CBT, PBT and CBT with 4DRT technology may be expected to improve LC and OS for larger tumors and for multiple tumors in parallel-type OAR, and small tumors in patients with medical complications who have previously not been candidates for curative radiotherapy.

4.2. Cancers in/near serial-type OAR.

In the late 1980s, pioneer physicians started high-precision 3DRT giving high-doses to small intra-cranial lesions in/near the brain, a serial OAR, using fixation devices for the cranium. Since organ motion of intra-cranial structures is very small, the 4DRT technique is not required for intra-cranial lesions. High-precision 3DXT giving high doses to small tumors has replaced conventional whole brain homogeneous irradiation for a few brain metastases avoiding neurological late toxicities.¹⁷³⁾^–¹⁷⁵⁾ The recommended dose at present is 18 Gy/1 Fr (EDQ2₂ = 90 Gy, EQD2₁₀ = 42 Gy) for metastatic lesions ≤10 mm in radius keeping the brain V_12Gy ≤ 5 cc (radius = 10.6 mm) of the brain.¹²⁹⁾^,¹⁷⁶⁾ The dose constraint for the brain, V_12Gy ≤ 5 cc, i.e., 12 Gy/1 Fr (EQD2₂ = 42 Gy) in high-precision 3DXT, is biologically the same as the dose constraint of conventional 42 Gy/21 Fr (EQD₂ = 42 Gy) in 2DXT. Even with high-precision 3DXT, dose fractionation is recommended for larger metastatic brain tumors suggesting that 18 Gy/1 Fr is less safe for the brain and less effective for the tumor if the high dose volume is larger than 5 cc.¹²⁹⁾^,¹⁷⁶⁾^,¹⁷⁷⁾ In short, the dose safely given for intra-cranial lesions in/near serial OAR, for the brain, can be increased somewhat by high-precision 3DXT but not as much as in parallel OAR even when the intra-cranial lesion is ≤10 mm in radius. Recommended dose constraints in spinal SRS are also conservative like the brain: 14 Gy/1 Fr (EQD2₂ = 56 Gy) to 0.03 cc (radius = 2 mm) of spinal cord, 10 Gy/1 Fr (EQD2₂ = 30 Gy) to absolute spinal cord volumes smaller than 0.35 cc (radius = 4 mm), and 10 Gy/1 Fr to no more than 10% of the partial spinal cord volume (5–6 mm above the target spine to 5–6 mm below the target spine).¹⁷⁸⁾^,¹⁷⁹⁾ In recent RCT protocols, the dose constraint to the spinal cord in high-precision 3DXT was 17 Gy/2 Fr (EQD2₂ = 44.6 Gy), which is biologically similar to the dose constraint of 44 Gy/22 Fr (EQD2₂ = 44 Gy) in conventional 2DXT.¹⁸⁰⁾ However, using the same dose constraints to the spinal cord, high-precision 3DXT was able to give 24 Gy/2 Fr (EQD2₂ = 84 Gy, EQD2₁₀ = 44 Gy) to vertebral metastatic tumors improving the therapeutic radio for pain at 3 months compared with the 2DXT, 20 Gy/5 Fr (EQD2₂ = 43.8 Gy, EQD2₁₀ = 31.3 Gy).¹⁸⁰⁾ It is notable that strict quality assurance for the dose distribution was ensured for high-precision 3DXT in this study.¹⁸⁰⁾ These results suggest that the maximal dose to the vertebral tumor near spinal cord, serial-type OAR, can be somewhat increased by high-precision 3DXT if the high dose volume is as small as a few mm in radius.

In summary, we have been able to reduce the maximal dose and thus radiation-induced adverse effects maintaining the same effect to the tumor in/near brain and spinal cord, serial-type OAR by precise 3DXT. The maximal dose to a part of a serial-type OAR has not been increased very much by decreasing the irradiated volume but there may be mild volume effects in the serial-type OAR at the very small volume level such as volumes with 1 to 20 mm in radius. It may be possible to increase the maximum dose to the serial-type OAR only when the volume which receives higher doses is as small as 1 to 20 mm in radius. It is not possible to expect volume effects for other serial-type OAR such as the esophagus, trachea, large vessels, and digestive tract larger than the central nervous system. Since the maximal dose of the extra-cranial serial-type OAR can change drastically with the organ motion, very precise 4DXT is expected to improve the clinical outcome of cancers in/near extra-cranial serial-type OAR compared to that in high-precision 3DXT.

Prostate cancer is the typical example of extra-cranial cancer located adjacent to serial-type OAR, the rectum. Therefore, 4DXT can be expected to improve the therapeutic ratio compared with high-precision 3DXT. Table 3 shows the results of a non-randomized comparison between high-precision 3DXT and 4DXT for prostate cancer within a prospective trial where the backgrounds of the patients were equivalent.¹⁸¹⁾^,¹⁸²⁾ The comparison between 1mm-3DXT (CBCT±m) and 0.1s-4DXT (m+30DF+p) indicates that the 4DXT produced fewer late adverse effects.¹⁸¹⁾ However, the margin for the tumor in the direction to the rectal wall was different for 3DXT and 4DXT in the protocol: 5 mm in high-precision 3DXT and 3 mm in 4DXT. Therefore, it is a real question whether the 4DXT technology or the smaller margin was the main reason for the fewer late adverse effects, a question that still remains to be answered. Table 4 shows the results of RCT comparing MRI-guided and CT-guided SBRT for prostate cancer. Using the nonproprietary name, the comparison was between 1mm-3DXT (CBCT+m) and 1s-4DXT (0.25sMRI).¹⁸³⁾ In the article, early genitourinary (GU) and gastrointestinal (GI) adverse events were significantly less frequent in 1s-4DXT (0.25sMRI) compared to 1mm-3DXT (CBCT+m).¹⁸³⁾ However, the margin for the rectal wall was 4 mm in high-precision 3DXT and 2 mm in 4DXT. Therefore, the question of whether 4DXT technology or the smaller margin was the basic reason of the fewer late adverse effects remains to be answered. Among 4DXT with the same margins, Table 5 shows comparisons in the literature between the early and late adverse events in 0.1s-4DXT (m+0.03DF) and in 0.1s-4DXT (m+30DF+p) for prostate cancer.¹⁸¹⁾^,¹⁸²⁾^,¹⁸⁴⁾^,¹⁸⁵⁾ The 0.1s-4DXT (m+0.03DF) may be somewhat better for early adverse events but there seems to be little difference in later effects. The apparent difference in early adverse events may be due to differences in dose fractionations and target volume settings by institutions. No comparison between 0.1s-4DXT (m+30DF+p) and 1s-4DXT (0.25sMRI) has yet been made. In summary, for prostate cancer which is inevitably near the rectum, serial OAR, 0.1s-4DXT (m+30sDF+p) and 1s-4DXT (0.25MRI) corresponding to the slow-type motion are likely to be superior to 1mm-3DXT (CBCT±m) to reduce adverse effects. However, the width of the margin for the rectum direction remains the commissure factor in this comparison. Tumor control and survival outcome data are awaited.

Table 3. Non-randomized comparison between high-precision 3DRT and 4DRT for prostate cancer in the stereotactic body radiotherapy arm in an open-label, randomized, phase 3, non-inferiority trial (PACE-B)¹⁸¹⁾^,¹⁸²⁾

Nonproprietary Name	Names in the Literature	PTV Margin for CTV (Rectal Direction)	Dose Fractionation (Gy/Fractions)	Number of Patients	Adverse Effects
Nonproprietary Name	Names in the Literature	PTV Margin for CTV (Rectal Direction)	Dose Fractionation (Gy/Fractions)	Number of Patients	Late (2-years)	Patient-reported QOL (2-years)
1mm-3DXT (CBCT±m)	conventional linac	5 mm	36.25 Gy/5 Fr	212	GU >G2 17% GI >G2 5%	GU score decreased 37%
0.1s-4DXT (m+30sDR+p)	robotic non- coplanar RT (CyberKnife^®)	3 mm	36.25 Gy/5 Fr	154	GU >G2 6% (p = 0.002) GI >G2 1% (p = 0.016)	GU score decreased 25% (p = 0.036)

PTV: planning target volume, CTV: clinical target volume, QOL: quality of life, 3DXT: three-dimensional x-ray therapy, CBCT: cone-beam computed tomography, GU: genitourinary, GI: gastrointestinal, 4DXT: four-dimensional x-ray therapy, DR: dual radiography, m: marker, p: prediction model, RT: radiotherapy.

Table 4. Randomized Comparison between high-precision 3DRT and 4DRT for Prostate Cancer in an Open-labeld, Randomized, Phase 3, Clinical Trial (MIRAGE) at a Single Center by Kishan et al. 2023¹⁸³⁾

Nonproprietary Name	Names in the Literature	PTV Margin for CTV (Rectal Direction)	Dose Fractionation (Gy/Fractions)	Number of Patients	Adverse Events
Nonproprietary Name	Names in the Literature	PTV Margin for CTV (Rectal Direction)	Dose Fractionation (Gy/Fractions)	Number of Patients	Early	Patient-reported IPSS
1mm-3DXT (CBCT+m+DR)	TrueBeam^® or Novalis Tx^®	4 mm	40 Gy/5 Fr	77	GU >G2 43.4% GI >G2 10.5%	>15 at 1 month 19.4% >15 at 3 months 1.4%
1s-4DXT (0.25sMRI)	MRIdian^®	2 mm	40 Gy/5 Fr	79	GU >G2 24.4% (p = 0.01) GI >G2 0.0% (p = 0.003)	>15 at 1 months 6.8% (p = 0.01) >15 at 3 months 4.1% (n.s.)

PTV: planning target volume, CTV: clinical target volume, IPSS: International Prostate Symptom Score, 3DXT: three-dimensional x-ray therapy, CBCT: cone-beam computed tomography, m: marker, DR: dual radiography, GU: genitourinary, GI: gastrointestinal, 4DXT: four-dimensional x-ray therapy.

Table 5. Comparison of early and late adverse effects between different 4DRT systems using the same PTV margin reported in literature

Nonproprietary Name	Names in the Literature	PTV Margin for CTV	Dose Fractionation (Gy/Fractions)	Number of Patients (follow-up years, yrs)	Adverse Events (Grade, G)		Remarks
Nonproprietary Name	Names in the Literature	PTV Margin for CTV	Dose Fractionation (Gy/Fractions)	Number of Patients (follow-up years, yrs)	Early	Late	Remarks
0.1s-4DXT (m+30sDR+p)₅	CyberKnife^®	3 mm	36.25 Gy/5 Fr (Urethra V47 <20%)	309 (Mean 5 yrs)	GU G2 26% G3 0% GI G2 8% G3 0%	GU G2 12% G3 1% GI G2 0% G3 0%	Meier et al. 2018¹⁸⁴⁾
0.1-4DXT (m+30sDR+p)₅	CyberKnife^®	3 mm	36.25 Gy/5 Fr	154 (93%>2-yrs)	GU G2 6% G3 0% GI G2 4% G3 0%	at 2-years GU G2 6% G3 0% GI G2 1% G3 0%	Brand et al. 2019¹⁸¹⁾ Tree et al. 2022¹⁸²⁾
0.1s-4DXT (m+0.03sDF)₃₀	Varian Linac +SyncTraX^®	3 mm	70 Gy/30 Fr PTV D95 (Urethra V70 <10%)	110 (Mean 2.5 yrs)	GU G2 1% G3 0% GI G2 0% G3 0%	GU G2 3% G3 0% GI G2 0% G3 0%	Shimizu S et al. 2014¹⁸⁵⁾

PTV: planning target volume, CTV: clinical target volume, 4DXT: four-dimensional x-ray therapy, Urethra V47: volume of urethra receiving more than 47 Gy, Urethra V70: volume of urethra receiving more than 70 Gy, GU: genitourinary, GI: gastrointestinal.

4.3. Cancers in complex regions.

Lung cancer at the central part of the lung is a typical example of a cancer at the complex region consisting of serial-type and parallel-type OAR. The OAR are large vessels, the trachea, main bronchus, and esophagus, all of which are serial OAR, in addition to the lungs, a parallel OAR. Onimaru et al. have reported that one patient in whom 1 cc of the esophagus received 42.5 Gy/8 Fr (D_1cc = 42.5 Gy/8 Fr, EQD2₂ = 77.7 Gy) with the D_max of 50.5 Gy/8 Fr (EQD2₂ = 104.9 Gy), died due to an esophageal ulcer after high-precision 3DXT for stage I NSCLC.¹⁸⁶⁾ Serious attention must be paid to the enthusiasm for high-precision 3DXT or 4DXT giving high doses to the centrally located stage I NSCLC near these serial OARs because radiation damage to these OARs is life threatening.¹⁸⁷⁾ It is prudent to use the same dose and fractionation as that used in 2DXT for the centrally located stage I NSCLC since central lung cancers arise from the proximal bronchial tree, also a serial OAR, and precision in the localization is not possible to differentiate the tumor from OAR. However, because surgery for centrally located stage I NSCLC is often very invasive, high-precision 3DXT giving an ablative dose has been investigated in prospective trials as an alternative to surgery. Timmerman et al. performed prospective study of high-precision 3DXT for central stage I NSCLC giving 60–66 Gy/3 Fr (EQD2₂ = 330–396 Gy, EDQ2₁₀ = 150–176 Gy) for tumors within 20 mm of the proximal bronchial tree, a serial OAR, and reported that the 2-year freedom from severe adverse events was only 54%.¹⁸⁸⁾ Kimura et al. have reported that 60 Gy/8 Fr (EQD2₂ = 142.5 Gy) was reaching the maximum tolerable dose for the centrally located stage I NSCNC because it is technically impossible either with high-precision 3DXT or 4DXT to reduce the dose to serial OAR.¹⁸⁹⁾ Recently, D_1cc < 40 Gy/8 Fr (EQD2₂ = 70 Gy) was recommended as the tolerance dose for the esophagus which was predicted to result in 1.1% high-grade esophageal adverse events for the centrally located stage I NSCLC.¹⁹⁰⁾

Other examples of the cancers in complex regions are cancers arising in parallel-type OAR but located near the abdominal digestive tract, such as cancers in the lower lobe of the left lung, the central or lower part of the liver, bile duct, pancreas, adrenal gland, and kidneys. Jang et al. showed that, in 65 patients with HCC, high-precision 3DXT or 4DXT was safe and effective in general but an esophageal ulcer with stenosis occurred at 5 months after giving D_max 42 Gy/3 Fr (EQD2₂ = 168 Gy) to the esophagus and the incidence of gastritis and duodenitis increased rapidly if the D_2cc of the gastroduodenum was ≥10 Gy/3 Fr (EQD2₂ = 13.3 Gy).¹⁹¹⁾ Gigantic HCC has tolerated high doses with 3D PBT¹⁷⁰⁾ but serious esophageal ulcers have developed after 3D PBT when the large tumor is close to the esophagus.¹⁹²⁾

Pancreatic cancer arises from a complex region consisting of serial-type OAR such as the stomach, duodenum, and other bowel structures, and parallel-type OAR such as the liver and spleen. The history of radiotherapy for pancreatic cancer has been a cycle of hope and disappointments.¹⁹³⁾ Table 6 shows clinical results of 3DXT and 4DXT for pancreatic cancers reported before 2020. A retrospective study showed neoadjuvant chemoradiotherapy using conventional 3DXT (body frame+DR) giving 50–60 Gy/25 Fr (EQD2₂ = 50–66 Gy, EQD2₁₀ = 50–62 Gy) followed by surgery resulted in 54.5% 5-years OS with 1.9% (3/157) late GI adverse events in 157 patients who underwent radical surgical resection for operable or borderline pancreatic cancers.¹⁹⁴⁾ However, in a RCT involving patients with locally advanced pancreatic cancers with the disease controlled after 4 months of induction chemotherapy, there was no significant difference in overall survival with chemoradiotherapy using 1mm-3DXT (DR) giving 54 Gy/30 Fr (EQD2₂ = 51.3 Gy, EQD2₁₀ = 53.1 Gy) compared with chemotherapy alone.¹⁹⁵⁾ A prospective phase II study of 1mm-3DXT (body frame+DR)₃ giving 45 Gy/3 Fr (EQD2₂ = 191.3 Gy, EQD2₁₀ = 93.8 Gy) for inoperable T1-3N0M0 pancreatic cancers showed a poor clinical outcome, 5% (0–12%) 1-year OS with a significant deterioration in performance status, increased nausea, and increased pain.¹⁹⁶⁾ Using 0.1s-4DXT (m+30sDF+p), giving 25 Gy/1 F (EQD2₂ = 168.8 Gy, EQD2₁₀ = 72.9 Gy), late Gr 3 or greater GI adverse events occurred in 12.5% (2/16) of the patients who survived >4 months after radiotherapy for locally advanced, non-metastatic pancreatic cancer in a prospective study (Table 6).¹⁹⁷⁾

Table 6. Clinical results of 3DRT and 4DRT for pancreatic cancer reported before 2020

Nonproprietary Name	Names in the Literature	PTV Margin for CTV	Radiotherapy (Gy/Fractions (Fr)) and Other Treatment*	Median (month (m)) Overall Survival (OS) (years (yrs)) from Diagnosis/ Registration	Adverse Events (Grade, (G))	Remarks
1mm-3DXT (body frame+DR)	A 10-MV linear accelerator	5 mm to CTV on exhale and inhale phases	50–60 Gy/25 Fr Concurrent chemotherapy followed by surgery	Median 74.5 m 5-yrs OS 54.5%	GI G3 1.9% (3/157)	Hirata et al. 2015¹⁹⁴⁾ 157 operated patients out of 222 patients with resectable or borderline resectable cancer.
1mm-3DXT (DR)	3D Conformal Radiotherapy	15 mm AP&LR 30 mm CC for GTV	54 Gy/30 Fr Induction and concurrent chemotherapy	Median 15.2 m [95% CI 13.9–17.3 m]	GI >G3 5.9% (6/133)	Hammel et al. 2016¹⁹⁵⁾ 133 patients with locally advanced cancer.
1mm-3DXT (body frame+DR)₃	Stereotactic Body Frame^® or vacuum pillow, Primus^® or Clinic 2100/2300^®	5 mm AP & LR 10 mm CC for CTV	45 Gy/3 Fr	Median^† 5.4 m 1-yr OS^† 5% [95% CI 0–12%]	GI >G2 at the start^‡ 64% at 14 days 79%	Hoyer et al. 2005¹⁹⁶⁾ 22 patients with inoperable locally advanced cancer.
0.1s-4DXT (m+30sDR+p)₁	CyberKnife^® (with Synchrony^®)	2–3 mm to CTV on exhale and inhale phases	25 Gy//1 Fr Induction and concurrent chemotherapy	Median 11.4 m 1-yr OS 50% 2-yrs OS 18%	GI >G3 12.5% (2/16)	Schellenberg et al. 2008¹⁹⁷⁾ 16 patients with locally advanced cancer.

PTV: planning target volume, CTV: clinical target volume, 3DXT: three-dimensional x-ray therapy, DR: dual radiography, CC: cranio-caudal, AP: antero-posterior, LR: left-right, GI: gastrointestinal, 4DXT: four-dimensional x-ray therapy, †: From the start of radiotherapy, ‡: Proportion of patients with the symptoms before the treatment. *All studies used adjuvant chemotherapy if possible.

Table 7 shows clinical results of 3DXT and 4DXT for pancreas cancer reported after 2020. Katz et al. have shown that 33–45 Gy/5 Fr (EQD2₂ = 71–123.8 Gy, EQD2₁₀ = 45.7–71.3 Gy) or 25 Gy/5 Fr (EQD2₂ = 43.8 Gy, EQD2₁₀ = 31.3 Gy) using 1mm-3DRT (m+DR) with induction chemotherapy and surgery of borderline resectable adenocarcinomas of the pancreas is well tolerated (>Gr 3 GI adverse events 7%) but without improvement in the survival in a randomized phase II trial.¹⁹⁸⁾ It is notable that 1s-4DXT (0.25sMRI)₅/IPP giving 40–50 Gy/5 Fr (EQD2₂ = 100–150 Gy, EQD2₁₀ = 60–83.3 Gy) followed by surgery for responders to preoperative radiotherapy with or without induction chemotherapy showed 82% 2-years OS.¹⁹⁹⁾ The 2-years OS for 62 pancreas cancers after induction chemotherapy followed by 1s-4DXT (0.25sMRI)₅/IPP giving 50 Gy/5 Fr was 45.5% which was better than chemotherapy alone in the literature.¹³⁰⁾ The benefit of 1s-4DXT (0.25sMRI)₅/IPP was investigated in prospective studies showing that 50 Gy/5 Fr after >3 months of induction chemotherapy achieved 93.9% 1-year OS with adverse events comparable with the risk of surgical resection without radiotherapy.²⁰⁰⁾ These encouraging results of 1s-4DXT (0.25sMRI₅)/IPP are consistent with the theoretical expectations for 4DXT. It is notable that in the dose constraints for GI structures, serial OAR, are all for volumes ≤ 0.5 cc (radium = 5 mm) in the protocol for 1s-4DXT (0.25sMRI)₅/IPP which fits the prediction from the biological considerations described above. With 1s-4DXT (0.25sMRI)₅/IPP for fast-type motion there may be increases in the role of radiotherapy for pancreatic cancer, although caution is needed about the critical serial-type OAR in the abdomen such as the major arteries and portal veins especially for cancers that require surgical reconstruction of vessels after radiotherapy. It is prudent to use the same dose and fractionation as used in 2DXT for these OAR if precision in localization is not possible to differentiate pancreatic cancer form OAR.

Table 7. Clinical Results of 3DRT and 4DRT for pancreatic cancer reported in literature

Nonproprietary Name	Names in the Literature	PTV Margin for CTV	Radiotherapy (Gy/Fractions (Fr)) and Other Treatment*	Median (month (m)) Overall Survival (OS) (years (yrs)) from Diagnosis/ Registration	Adverse Events (Grade, (G))	Remarks
1mm-3DXT (m+DR)	SBRT or high-dose IGRT	3 mm in SBRT 5–10 mm in IGRT	SBRT 33–40 Gy/ 5 Fr or IGRT 25 Gy/5 Fr Induction chemotherapy & surgery for responders	Median 17.1 m [95% CI: 12.8–24.4 m] 1.5-yrs OS 47.3% [95% CI: 35.8–62.5%]	Early GI G3 7% (3/44) After surgery >G3 47% (9/19)	Katz et al. 2022¹⁹⁸⁾ 55 patients with borderline resectable cancer. 44 irradiated and 19 resected.
1s-4DXT (0.25sMRI)₅/IPP	MRIdian^®	3 mm for CTV (5 mm for OAR) Isotoxic plan policy	40–50 Gy/5 Fr ± Induction chemotherapy & surgery for responders	Median Not reached 1-yr OS 100% 2-yrs OS 82% [95% CI: 61–93%]	Early GI >G2 0% After surgery >G3 16% (4/26) G5 0%	Bryant et al. 2022¹⁹⁹⁾ 26 patients with localized cancer irradiated and received resection with curative intent.
1s-4DXT (0.25sMRI)₅/IPP	MRIdian^®	3 mm for CTV (3–5 mm for OAR) Isotoxic plan policy	40–50 Gy/5 Fr Induction chemotherapy & surgery for responders	Median 23 m [95% CI: 18.0–29.0 m] 1-yr OS 90.2% 2-yrs OS 45.5% [95% CI: 31.5–59.5%]	Early GI G3 4.8% Late GI G3 4.8% After surgery G5 14.3% (2/14)	Chuong et al. 2022¹³⁰⁾ 62 patients with locally advanced cancer irradiated. 14 patients resected after radiotherapy.
1s-4DXT (0.25sMRI)₅/IPP	0.35T MR- ⁶⁰Cobalt or MR-linac system	3 mm for CTV Isotoxic plan policy	50 Gy/5 Fr Induction chemotherapy & surgery for responders	1-yr OS 93.9%	Early GI >G3 8.8% (12/136) After surgery >G3 20.5% (9/44) G5 4.5% (2/44)	Parikh et al. 2023²⁰⁰⁾ 136 patients with locally advanced cancer irradiated. 44 patients resected after radiotherapy.

PTV: planning target volume, CTV: clinical target volume, 3DXT: three-dimensional x-ray therapy, m: marker, DR: dual radiography, SBRT: stereotactic body radiotherapy, IGRT: image-guided radiotherapy, 4DXT: four-dimensional x-ray therapy, MRI: magnetic resonance imaging, IPP: isotoxic planning policy, OAR: organ at risk, IGRT: image-guided radiotherapy, GI: gastrointestinal. *All studies used adjuvant chemotherapy if possible.

In summary, usage of nonproprietary names of radiotherapy devices has highlighted the challenges of current radiotherapy. Future clinical trials of radiotherapy for cancers in complex regions consisting of serial- and parallel-type OAR, such as pancreatic cancers, should be able to accurately assess doses to serial OAR. New development in 4DRT planning and delivery systems based on spatio-temporal co-ordinates are expected to pave the way here. Clinical trials of 4DRT that do not exclude cancers near serial OAR can be expected to be conducted. However, for any new developments, it must be physically proven in advance that the 4DRT technique can control the dose limit for the part of the serial-type OAR from approximately 1 to 20 mm in radius with sufficient doses to the moving tumor. Where this is not possible, it is important to prescribe conventional doses and multiple fractionations with certified safety from the long history of 2DRT.

5. Limitation of this review and future prospects

This review has several limitations. First, the review about NTCP and TCP models is not comprehensive and written from limited viewpoints for simplicity. Clinically, we should seriously take information such as the Quantitative Analyses of Normal Tissue Effects in the Clinic (QUANTEC)¹²⁴⁾ and Hypofractionated Treatment Effects in the Clinic (HyTEC)¹²⁵⁾ which are carefully detailed not to be pedantic about the biophysical models for volume effect and showing dose/volume/outcome data based on enormous clinical data volumes. Secondly, motion management technology specific to particle beam therapy such as repainting and 4D robust optimization is not covered in this review. It will be worthwhile to compare these among 4DXT, 4D PBT, and 4D CBT from biomedical aspects in the future.

We have seen that the physical accuracy, or higher dimensionality, in EBRT caused improvements in LC and OS in patients with peripheral small tumors in parallel-type OAR but may not provide advantages in/near serial-type OAR and in complex regions as reported so far. Apart from combinations with other therapies, which is outside the scope of this review, what can be posited as the guiding principle to improve EBRT further? Thinking about EBRT itself in spatio-temporal coordinates may give hints for the future.

Mizuta et al. have been the first to show that optimal dose fractionation in time changes with the physical dose distribution in space in EBRT.²⁰¹⁾ Multiple dose fractionations have made the therapeutic window wider in 2DRT which fits with the theoretical considerations using the linear-quadratic (LQ) model for cell survival curve.¹¹²⁾ The LQ model is now almost universally used for calculating isoeffect doses for different fractionation schedules in EBRT.²⁰²⁾ The concept of hyperfractionation was supported by predictions based on the LQ model and found to be useful for head and neck cancers.²⁰³⁾ According to the Mizuta et al. suggestion, however, hypofractionation using high-precision 3DRT and 4DRT is more suitable for tumors in parallel-type OAR to reduce the damage to the serial-type OAR near the tumor.²⁰¹⁾ The optimal number of fractions is determined by the physical 3D dose distribution and the α/β ratio of tumor and normal tissue in the LQ model.²⁰¹⁾ As we have seen above, damage to the serial-type OAR is the determinant for many types of tumors now. If the following holds, the damage to serial-type OAR near the tumor will be reduced, not increased, with hypofractionated irradiation compared with mutlfractionated irradiation:

\begin{equation} \frac{\alpha_{0}}{\beta_{0}}\biggm/\frac{\alpha_{1}}{\beta_{1}} \geq \delta \end{equation}

[2]

where α₀/β₀ is for serial-type OAR and α₁/β₁ is for cancer and δ is the ratio of D_max of OAR to the D_min of tumors, in principle. Since α₀/β₀ is usually 2–3 for late adverse effects of serial-type OAR, α₁/β₁ is 5–10 for cancer, and δ for serial-type OAR, such as for the main bronchus, trachea and esophagus, is less than 1 for peripheral lung cancers in high-precision 3DRT and 4DRT, hypofractionation is better for peripheral lung cancer. Still, if we treat centrally located lung cancer, where δ is almost always 1.0 irrespective EBRT technology, this equation cannot be maintained with just any EBRT technology making conventional doses and multiple fractionations are better here. This prediction is consistent with the consensus of clinical practice¹⁵⁸⁾^–¹⁶⁴⁾ and the results of clinical trials as reviewed above.¹⁸⁷⁾^,¹⁸⁸⁾ The next question then is what total dose and dose fractionation schedule should be used if a primary cancer at the peripheral lung and lymph nodes in the central part of the lung that need to be irradiated, and if we want to minimize the damage to the serial-type OAR. Using the same model and concept, simulation suggested that patients are better off when treated with hypofractionation to the primary NSLCL followed by multiple fractionations to the central metastatic lymphnodes (submitted for publication). The question becomes which tumor to treat first? Considering experimental results in small animals²⁰⁴⁾ and the immune activation cycle,²⁰⁵⁾^,²⁰⁶⁾ it may be that primary tumors should be treated first before affecting the draining lymph nodes, although there is no clinical evidence yet for this. This kind of investigation in the future would improve the therapeutic ratio of present 4DRT further. The limitations of ultra-high dose-rate pulsed therapy²⁰⁷⁾ may be overcome by introducing similar spatio-temporal 4D fractionation. The guiding principle of 4DRT in the future may become “Use Therapeutic Open Doors in Spatio-Temporal Co-ordinates”.

To realize the 4D fractionation, meticulous 4D precision at every point in the body, which is not available at this moment, will be required to minimize underdosage and overdosage at the boundary of each of the treatment volumes. The benefit of the high-precision 4DRT will be improved by the introduction of technology where the 3D coordinates around the cancer in the body at the moment of irradiation can be identified within ±1 mm, and irradiation can be performed within a delay time of 0.1 second. Planning systems in 4D coordinates, real-time verification of 3D dose distribution, automatic extraction of tumors and OAR, automatic immediate adaptive irradiation, minimally invasive fiducial markers for tumor near serial OAR, optimal 4D fractionation and immune activation simulators are subjects to be investigated further. These technologies would make high-precision 4DRT as the standard EBRT for complex regions and improve the clinical outcome for the patients. We are living in an era not far from the next changes to leap over/across this hurdle.

6. Conclusions

The question of “whether we would be able to increase the dose to the cancer keeping NTCP of the OAR below their thresholds by reducing the irradiated volume of the normal tissue” or “whether volume effect arises in normal tissue or not” depend on the location of the tumor and structure of the OAR surrounding the tumor. The TCP and NTCP can be controlled better in high-precision 3DRT than in 2DRT since the corresponding dosimetric indices of the tumor and the OAR can be controlled intentionally. Doses to cancers in a parallel-type OAR can be increased if the mean dose of the parallel-type OAR is below the threshold dose using high-precision 3DRT. Localization precision required for the cancers in parallel-type OAR is not so strict as that required for cancers in/near serial-type OAR and can be mitigated if the mean dose of the parallel-type OAR is much lower than the threshold dose. The dose to cancers in/near a serial-type OAR may be able to accommodate increased safely only if the volume of the serial-type OAR which is irradiated with a higher dose than conventional EBRT is smaller than a threshold volume, with approximately 1 mm to 20 mm in radius, depending on the OAR. It is often difficult to irradiate a tumor accurately because of organ motion. Precise irradiation at the millimeter level throughout the delivery of irradiation using 4DRT technology is increasingly used for moving tumors in/near extracranial serial-type OAR. Doses to cancers in a complex region consisting of serial- and parallel-type OAR can be increased by combining particle beam therapy and 4DRT technology.

Numerous clinical trials have been conducted about the biomedical safety and effectiveness of high-precision 3DRT and 4DRT. However, here various EBRT systems were equated under the name of high-precision EBRT in many clinical studies and EBRT systems were not included in the stratification criteria in most of RCT. Even when the EBRT systems were stratified before randomization, the protocol was such that the margin around the tumor can be changed for each EBRT system. Therefore, at this time it is not clear whether the cause of the improvement in safety by 4DRT for cancers in/near serial-type OAR is the difference brought about by the EBRT system or the difference in the margin around the tumor. Tumor control and survival outcome data are awaited.

Thinking about radiotherapy in spatio-temporal coordinates provides ideas for future prospects. Theoretically, the optimal dose fractionation in time changes with the physical 3D dose distribution in space. High-precision radiotherapy is a feat that will remain in the cultural history of mankind and will continue to develop in the future.

Acknowledgements

The author wishes to express appreciation for all the supporters of this review. This work is supported by KAKENHI (19H03591).

Notes

Edited by Tadao KAKIZOE, M.J.A.

Correspondence should be addressed to: H. Shirato, Global Center for Biomedical Science and Engineering, Faculty of Medicine, Hokkaido University, North-15 West-7, Kita-ku, Sapporo, Hokkaido 060-8638, Japan (e-mail: shirato@med.hokudai.ac.jp).

References

1) Röntgen, W.C. (1898) Ueber eine neue Art von Strahlen. Ann. Phys. 300, 12–17.
2) Coutard, H. (1934) Principles of x ray therapy of malignant diseases. Lancet 2, 1–8.
3) Paterson, R. (1939) Cancer: A consideration of curability and of the role of radiotherapy in treatment. Br. Med. J. 4, 904–907.
4) Umegaki, Y. (1957) The technics and results of the radiation therapy of oesophageal cancer. Gan 48, 512–513.
5) Takahashi, S. (1965) Conformation radiotherapy. Rotation techniques as applied to radiography and radiotherapy of cancer. Acta Radiol. Diagn. (Stockh.) Suppl 242, 241+.
6) Hounsfield, G.N. (1973) Computerized transverse axial scanning (tomography). 1. Description of system. Br. J. Radiol. 46, 1016–1022.
7) Chernak, E.S., Rodriguez-Antunez, A., Jelden, G.L., Dhaliwal, R.S. and Lavik, P.S. (1975) The use of computed tomography for radiation therapy treatment planning. Radiology 117, 613–614.
8) Tsujii, H., Bagshaw, M.A., Smith, A.R., von Essen, C.F., Mettler, F.A. and Kligerman, M.M. (1980) Localization of structures for pion radiotherapy by computerized tomography and orthodiagraphic projection. Int. J. Radiat. Oncol. Biol. Phys. 6, 319–325.
9) Mohan, R., Barest, G., Brewster, L.J., Chui, C.S., Kutcher, G.J., Laughlin, J.S. et al. (1988) A comprehensive three-dimensional radiation treatment planning system. Int. J. Radiat. Oncol. Biol. Phys. 15, 481–495.
10) ICRU (1993) ICRU Report 50, Prescribing, Recording, and Reporting Photon Beam Therapy. Bethesda, MD, USA.
11) ICRU (1999) ICRU Report 62, Prescribing, Recording and Reporting Photon Beam Therapy (Supplement to ICRU 50). Bethesda, MD, USA.
12) Brahme, A. (1987) Design principles and clinical possibilities with a new generation of radiation therapy equipment. A review. Acta Oncol. 26, 403–412.
13) Intensity Modulated Radiation Therapy Collaborative Working Group (2001) Intensity-modulated radiotherapy: current status and issues of interest. Int. J. Radiat. Oncol. Biol. Phys. 51, 880–914.
14) Shirato, H., Shimizu, S., Kitamura, K., Nishioka, T., Kagei, K., Hashimoto, S. et al. (2000) Four-dimensional treatment planning and fluoroscopic real-time tumor tracking radiotherapy for moving tumor. Int. J. Radiat. Oncol. Biol. Phys. 48, 435–442.
15) Shimizu, S., Shirato, H., Aoyama, H., Hashimoto, S., Nishioka, T., Yamazaki, A. et al. (2000) High-speed magnetic resonance imaging for four-dimensional treatment planning of conformal radiotherapy of moving body tumors. Int. J. Radiat. Oncol. Biol. Phys. 48, 471–474.
16) Baumann, M., Krause, M., Overgaard, J., Debus, J., Bentzen, S.M., Daartz, J. et al. (2016) Radiation oncology in the era of precision medicine. Nat. Rev. Cancer 16, 234–249.
17) Endo, M. (2022) Creation, evolution, and future challenges of ion beam therapy from a medical physicist’s viewpoint (part 1). Introduction and Chapter 1. accelerator and beam delivery system. Radiol. Phys. Technol. 15, 271–290.
18) Tsujii, H., Inada, T., Maruhashi, A., Hayakawa, Y., Tsuji, H., Ohara, K. et al. (1989) [Field localization and verification system for proton beam radiotherapy in deep-seated tumors]. Nihon Igaku Hoshasen Gakkai Zasshi 49, 622–629.
19) Endo, M. (2023) Creation, evolution, and future challenges of ion beam therapy from a medical physicist’s viewpoint (Part 2). Chapter 2. Biophysical model, treatment planning system and image guided radiotherapy. Radiol. Phys. Technol. 16, 137–159.
20) Vedam, S.S., Keall, P.J., Kini, V.R., Mostafavi, H., Shukla, H.P. and Mohan, R. (2003) Acquiring a four-dimensional computed tomography dataset using an external respiratory signal. Phys. Med. Biol. 48, 45–62.
21) Barendsen, G.W. (1982) Dose fractionation, dose rate and iso-effect relationships for normal tissue responses. Int. J. Radiat. Oncol. Biol. Phys. 8, 1981–1997.
22) Holthusen, H. (1936) Erfahrungen uber die vertraglichkeitsgrenze fur roentgenstrahlen und deren nutzanwendung zur verhutung von shaden. Strahlenther. Onkol. 57, 254–269.
23) Mackie, T.R., Holmes, T., Swerdloff, S., Reckwerdt, P., Deasy, J.O., Yang, J. et al. (1993) Tomotherapy: a new concept for the delivery of dynamic conformal radiotherapy. Med. Phys. 20, 1709–1719.
24) Jaffray, D.A., Siewerdsen, J.H., Wong, J.W. and Martinez, A.A. (2002) Flat-panel cone-beam computed tomography for image-guided radiation therapy. Int. J. Radiat. Oncol. Biol. Phys. 53, 1337–1349.
25) Verellen, D., De Ridder, M., Linthout, N., Tournel, K., Soete, G. and Storme, G. (2007) Innovations in image-guided radiotherapy. Nat. Rev. Cancer 7, 949–960.
26) Leksell, L. (1951) The stereotaxic method and radiosurgery of the brain. Acta Chir. Scand. 102, 316–319.
27) Shimizu, S., Shirato, H., Kagei, K., Nishioka, T., Bo, X., Dosaka-Akita, H. et al. (2000) Impact of respiratory movement on the computed tomographic images of small lung tumors in three-dimensional (3D) radiotherapy. Int. J. Radiat. Oncol. Biol. Phys. 46, 1127–1133.
28) van Herk, M., Remeijer, P., Rasch, C. and Lebesque, J.V. (2000) The probability of correct target dosage: dose-population histograms for deriving treatment margins in radiotherapy. Int. J. Radiat. Oncol. Biol. Phys. 47, 1121–1135.
29) Shimizu, S., Shirato, H., Xo, B., Kagei, K., Nishioka, T., Hashimoto, S. et al. (1999) Three-dimensional movement of a liver tumor detected by high-speed magnetic resonance imaging. Radiother. Oncol. 50, 367–370.
30) Shirato, H., Shimizu, S., Shimizu, T., Nishioka, T. and Miyasaka, K. (1999) Real-time tumour-tracking radiotherapy. Lancet 353, 1331–1332.
31) Shirato, H., Shimizu, S., Kunieda, T., Kitamura, K., van Herk, M., Kagei, K. et al. (2000) Physical aspects of a real-time tumor-tracking system for gated radiotherapy. Int. J. Radiat. Oncol. Biol. Phys. 48, 1187–1195.
32) Keall, P.J., Todor, A.D., Vedam, S.S., Bartee, C.L., Siebers, J.V., Kini, V.R. et al. (2004) On the use of EPID-based implanted marker tracking for 4D radiotherapy. Med. Phys. 31, 3492–3499.
33) Keall, P. (2004) 4-dimensional computed tomography imaging and treatment planning. Semin. Radiat. Oncol. 14, 81–90.
34) Ohara, K., Okumura, T., Akisada, M., Inada, T., Mori, T., Yokota, H. et al. (1989) Irradiation synchronized with respiration gate. Int. J. Radiat. Oncol. Biol. Phys. 17, 853–857.
35) Kubo, H.D. and Hill, B.C. (1996) Respiration gated radiotherapy treatment: a technical study. Phys. Med. Biol. 41, 83–91.
36) Lam, K.L., Ten Haken, R.K., McShan, D.L. and Thornton, A.F. (1993) Automated determination of patient setup errors in radiation therapy using spherical radio-opaque markers. Med. Phys. 20, 1145–1152.
37) Keall, P.J., Nguyen, D.T., O’Brien, R., Zhang, P., Happersett, L., Bertholet, J. et al. (2018) Review of real-time 3-dimensional image guided radiation therapy on standard-equipped cancer radiation therapy systems: Are we at the tipping point for the era of real-time radiation therapy? Int. J. Radiat. Oncol. Biol. Phys. 102, 922–931.
38) Wei, Z. (2022) AI-augmented image guidance for radiation therapy delivery. In Artificial Intelligence in Radiation Therapy (ed. Sumida, I.). Institute of Physics and Engineering in Medicine, Bristol, UK, pp. 1–24.
39) Shimizu, S., Shirato, H., Ogura, S., Akita-Dosaka, H., Kitamura, K., Nishioka, T. et al. (2001) Detection of lung tumor movement in real-time tumor-tracking radiotherapy. Int. J. Radiat. Oncol. Biol. Phys. 51, 304–310.
40) Engelsman, M., Sharp, G.C., Bortfeld, T., Onimaru, R. and Shirato, H. (2005) How much margin reduction is possible through gating or breath hold? Phys. Med. Biol. 50, 477–490.
41) Sharp, G.C., Jiang, S.B., Shimizu, S. and Shirato, H. (2004) Prediction of respiratory tumour motion for real-time image-guided radiotherapy. Phys. Med. Biol. 49, 425–440.
42) Wu, H., Sharp, G.C., Zhao, Q., Shirato, H. and Jiang, S.B. (2007) Statistical analysis and correlation discovery of tumor respiratory motion. Phys. Med. Biol. 52, 4761–4774.
43) Berbeco, R.I., Nishioka, S., Shirato, H., Chen, G.T. and Jiang, S.B. (2005) Residual motion of lung tumours in gated radiotherapy with external respiratory surrogates. Phys. Med. Biol. 50, 3655–3667.
44) Kanoulas, E., Aslam, J.A., Sharp, G.C., Berbeco, R.I., Nishioka, S., Shirato, H. et al. (2007) Derivation of the tumor position from external respiratory surrogates with periodical updating of the internal/external correlation. Phys. Med. Biol. 52, 5443–5456.
45) Ionascu, D., Jiang, S.B., Nishioka, S., Shirato, H. and Berbeco, R.I. (2007) Internal-external correlation investigations of respiratory induced motion of lung tumors. Med. Phys. 34, 3893–3903.
46) Seppenwoolde, Y., Berbeco, R.I., Nishioka, S., Shirato, H. and Heijmen, B. (2007) Accuracy of tumor motion compensation algorithm from a robotic respiratory tracking system: a simulation study. Med. Phys. 34, 2774–2784.
47) Harada, T., Shirato, H., Ogura, S., Oizumi, S., Yamazaki, K., Shimizu, S. et al. (2002) Real-time tumor-tracking radiation therapy for lung carcinoma by the aid of insertion of a gold marker using bronchofiberscopy. Cancer 95, 1720–1727.
48) Shimizu, S., Shirato, H., Kitamura, K., Shinohara, N., Harabayashi, T., Tsukamoto, T. et al. (2000) Use of an implanted marker and real-time tracking of the marker for the positioning of prostate and bladder cancers. Int. J. Radiat. Oncol. Biol. Phys. 48, 1591–1597.
49) Shirato, H., Oita, M., Fujita, K., Watanabe, Y. and Miyasaka, K. (2004) Feasibility of synchronization of real-time tumor-tracking radiotherapy and intensity-modulated radiotherapy from viewpoint of excessive dose from fluoroscopy. Int. J. Radiat. Oncol. Biol. Phys. 60, 335–341.
50) Shiinoki, T., Kawamura, S., Uehara, T., Yuasa, Y., Fujimoto, K., Koike, M. et al. (2016) Evaluation of a combined respiratory-gating system comprising the TrueBeam linear accelerator and a new real-time tumor-tracking radiotherapy system: a preliminary study. J. Appl. Clin. Med. Phys. 17, 202–213.
51) Shiinoki, T., Onizuka, R., Kawahara, D., Suzuki, T., Yuasa, Y., Fujimoto, K. et al. (2018) Estimation of patient-specific imaging dose for real-time tumour monitoring in lung patients during respiratory-gated radiotherapy. Phys. Med. Biol. 63, 065016.
52) Shirato, H., Suzuki, K., Sharp, G.C., Fujita, K., Onimaru, R., Fujino, M. et al. (2006) Speed and amplitude of lung tumor motion precisely detected in four-dimensional setup and in real-time tumor-tracking radiotherapy. Int. J. Radiat. Oncol. Biol. Phys. 64, 1229–1236.
53) Zhang, J., Wang, L., Li, X., Huang, M. and Xu, B. (2021) Quantification of intrafraction and interfraction tumor motion amplitude and prediction error for different liver tumor trajectories in cyberknife synchrony tracking. Int. J. Radiat. Oncol. Biol. Phys. 109, 1588–1605.
54) Hoogeman, M., Prévost, J.B., Nuyttens, J., Pöll, J., Levendag, P. and Heijmen, B. (2009) Clinical accuracy of the respiratory tumor tracking system of the cyberknife: assessment by analysis of log files. Int. J. Radiat. Oncol. Biol. Phys. 74, 297–303.
55) Iizuka, Y., Matsuo, Y., Ishihara, Y., Akimoto, M., Tanabe, H., Takayama, K. et al. (2015) Dynamic tumor-tracking radiotherapy with real-time monitoring for liver tumors using a gimbal mounted linac. Radiother. Oncol. 117, 496–500.
56) Kamino, Y., Takayama, K., Kokubo, M., Narita, Y., Hirai, E., Kawawda, N. et al. (2006) Development of a four-dimensional image-guided radiotherapy system with a gimbaled X-ray head. Int. J. Radiat. Oncol. Biol. Phys. 66, 271–278.
57) Schnarr, E., Beneke, M., Casey, D., Chao, E., Chappelow, J., Cox, A. et al. (2018) Feasibility of real-time motion management with helical tomotherapy. Med. Phys. 45, 1329–1337.
58) Akimoto, M., Nakamura, M., Mukumoto, N., Tanabe, H., Yamada, M., Matsuo, Y. et al. (2013) Predictive uncertainty in infrared marker-based dynamic tumor tracking with Vero4DRT. Med. Phys. 40, 091705.
59) Poulsen, P.R., Cho, B., Langen, K., Kupelian, P. and Keall, P.J. (2008) Three-dimensional prostate position estimation with a single x-ray imager utilizing the spatial probability density. Phys. Med. Biol. 53, 4331–4353.
60) Balter, J.M., Wright, J.N., Newell, L.J., Friemel, B., Dimmer, S., Cheng, Y. et al. (2005) Accuracy of a wireless localization system for radiotherapy. Int. J. Radiat. Oncol. Biol. Phys. 61, 933–937.
61) Harris, W., Yorke, E., Li, H., Czmielewski, C., Chawla, M., Lee, R.P. et al. (2022) Can bronchoscopically implanted anchored electromagnetic transponders be used to monitor tumor position and lung inflation during deep inspiration breath-hold lung radiotherapy? Med. Phys. 49, 2621–2630.
62) Smith, R.L., Lechleiter, K., Malinowski, K., Shepard, D.M., Housley, D.J., Afghan, M. et al. (2009) Evaluation of linear accelerator gating with real-time electromagnetic tracking. Int. J. Radiat. Oncol. Biol. Phys. 74, 920–927.
63) Grimwood, A., McNair, H.A., O’Shea, T.P., Gilroy, S., Thomas, K., Bamber, J.C. et al. (2018) In vivo validation of Elekta’s clarity autoscan for ultrasound-based intrafraction motion estimation of the prostate during radiation therapy. Int. J. Radiat. Oncol. Biol. Phys. 102, 912–921.
64) Zhang, W., Oraiqat, I., Litzenberg, D., Chang, K.W., Hadley, S., Sunbul, N.B. et al. (2023) Real-time, volumetric imaging of radiation dose delivery deep into the liver during cancer treatment. Nat. Biotechnol. 41, 1160–1167.
65) Saito, M., Sano, N., Ueda, K., Shibata, Y., Kuriyama, K., Komiyama, T. et al. (2018) Technical Note: Evaluation of the latency and the beam characteristics of a respiratory gating system using an Elekta linear accelerator and a respiratory indicator device, Abches. Med. Phys. 45, 74–80.
66) Otani, Y., Fukuda, I., Tsukamoto, N., Kumazaki, Y., Sekine, H., Imabayashi, E. et al. (2010) A comparison of the respiratory signals acquired by different respiratory monitoring systems used in respiratory gated radiotherapy. Med. Phys. 37, 6178–6186.
67) Woods, K. and Rong, Y. (2015) Technical Report: TG-142 compliant and comprehensive quality assurance tests for respiratory gating. Med. Phys. 42, 6488–6497.
68) Murphy, M.J. (2004) Tracking moving organs in real time. Semin. Radiat. Oncol. 14, 91–100.
69) Pepin, E.W., Wu, H. and Shirato, H. (2011) Dynamic gating window for compensation of baseline shift in respiratory-gated radiation therapy. Med. Phys. 38, 1912–1918.
70) Chen, G.P., Tai, A., Puckett, L., Gore, E., Lim, S., Keiper, T. et al. (2021) Clinical implementation and initial experience of real-time motion tracking with jaws and multileaf collimator during helical tomotherapy delivery. Pract. Radiat. Oncol. 11, e486–e495.
71) Yamazaki, R., Nishioka, S., Date, H., Shirato, H., Koike, T. and Nishioka, T. (2012) Investigation of the change in marker geometry during respiration motion: a preliminary study for dynamic-multi-leaf real-time tumor tracking. Radiat. Oncol. 7, 218.
72) Harada, K., Katoh, N., Suzuki, R., Ito, Y.M., Shimizu, S., Onimaru, R. et al. (2016) Evaluation of the motion of lung tumors during stereotactic body radiation therapy (SBRT) with four-dimensional computed tomography (4DCT) using real-time tumor-tracking radiotherapy system (RTRT). Phys. Med. 32, 305–311.
73) Khadige, M., Salleron, J., Marchesi, V., Oldrini, G., Peiffert, D. and Beckendorf, V. (2018) Cyberknife. J. Thorac. Dis. 10, 4976–4984.
74) Imura, M., Yamazaki, K., Shirato, H., Onimaru, R., Fujino, M., Shimizu, S. et al. (2005) Insertion and fixation of fiducial markers for setup and tracking of lung tumors in radiotherapy. Int. J. Radiat. Oncol. Biol. Phys. 63, 1442–1447.
75) Moran, D.E., Parikh, M., Sheiman, R.G., Brook, O.R., Sun, M.R.M., Mahadevan, A. et al. (2021) Comparison of technical success and safety of transbronchial versus percutaneous CT-guided fiducial placement for SBRT of lung tumors. J. Med. Imaging Radiat. Sci. 52, 409–416.
76) Kitamura, K., Shirato, H., Shimizu, S., Shinohara, N., Harabayashi, T., Shimizu, T. et al. (2002) Registration accuracy and possible migration of internal fiducial gold marker implanted in prostate and liver treated with real-time tumor-tracking radiation therapy (RTRT). Radiother. Oncol. 62, 275–281.
77) Steinfort, D.P., Siva, S., Kron, T., Chee, R.R., Ruben, J.D., Ball, D.L. et al. (2015) Multimodality guidance for accurate bronchoscopic insertion of fiducial markers. J. Thorac. Oncol. 10, 324–330.
78) Ikeda, K., Liu, H., Miyamoto, N., Nguyen, M.T., Shirato, H. and Yonezawa, T. (2022) Preparation of biopex-supported gold nanoparticles as potential fiducial markers for image-guided radiation therapy. ACS Appl. Bio. Mater. 5, 1259–1266.
79) Liu, H., Miyamoto, N., Nguyen, M.T., Shirato, H. and Yonezawa, T. (2022) Injectable fiducial marker for image-guided radiation therapy based on gold nanoparticles and a body temperature-activated gel-forming system. ACS Appl. Bio. Mater.
80) Mueller, M., Poulsen, P., Hansen, R., Verbakel, W., Berbeco, R., Ferguson, D. et al. (2022) The markerless lung target tracking AAPM Grand Challenge (MATCH) results. Med. Phys. 49, 1161–1180.
81) Lagendijk, J.J., Raaymakers, B.W. and van Vulpen, M. (2014) The magnetic resonance imaging-linac system. Semin. Radiat. Oncol. 24, 207–209.
82) Fischer-Valuck, B.W., Henke, L., Green, O., Kashani, R., Acharya, S., Bradley, J.D. et al. (2017) Two-and-a-half-year clinical experience with the world’s first magnetic resonance image guided radiation therapy system. Adv. Radiat. Oncol. 2, 485–493.
83) Jassar, H., Tai, A., Chen, X., Keiper, T.D., Paulson, E., Lathuilière, F. et al. (2023) Real-time motion monitoring using orthogonal cine MRI during MR-guided adaptive radiation therapy for abdominal tumors on 1.5T MR-Linac. Med. Phys. 50, 3103–3116.
84) Kim, T., Lewis, B., Lotey, R., Barberi, E. and Green, O. (2021) Clinical experience of MRI^4D QUASAR motion phantom for latency measurements in 0.35T MR-LINAC. J. Appl. Clin. Med. Phys. 22, 128–136.
85) Boldrini, L., Alongi, F., Romano, A., Charles Davies, D., Bassetti, M., Chiloiro, G. et al. (2023) Current practices and perspectives on the integration of contrast agents in MRI-guided radiation therapy clinical practice: A worldwide survey. Clin. Transl. Radiat. Oncol. 40, 100615.
86) Otazo, R., Lambin, P., Pignol, J.P., Ladd, M.E., Schlemmer, H.P., Baumann, M. et al. (2021) MRI-guided radiation therapy: An emerging paradigm in adaptive radiation oncology. Radiology 298, 248–260.
87) Akdag, O., Borman, P.T.S., Woodhead, P., Uijtewaal, P., Mandija, S., Van Asselen, B. et al. (2022) First experimental exploration of real-time cardiorespiratory motion management for future stereotactic arrhythmia radioablation treatments on the MR-linac. Phys. Med. Biol. 67, 065003.
88) Shortall, J., Vasquez Osorio, E., Cree, A., Song, Y., Dubec, M., Chuter, R. et al. (2021) Inter- and intra-fractional stability of rectal gas in pelvic cancer patients during MRIgRT. Med. Phys. 48, 414–426.
89) Borman, P.T.S., Tijssen, R.H.N., Bos, C., Moonen, C.T.W., Raaymakers, B.W. and Glitzner, M. (2018) Characterization of imaging latency for real-time MRI-guided radiotherapy. Phys. Med. Biol. 63, 155023.
90) Wang, Y., Cai, B., Conroy, L. and Qi, X.S. (2022) Artificial intelligence in adaptive radiation therapy. In Artificial Intelligence in Radiation Therapy (ed. Sumida, I.). Institute of Physics and Engineering in Medicine, Bristol, UK, pp. 1–20.
91) Li, H., Dong, L., Bert, C., Chang, J., Flampouri, S., Jee, K.W. et al. (2022) AAPM Task Group Report 290: Respiratory motion management for particle therapy. Med. Phys. 49, e50–e81.
92) Matsuura, T., Miyamoto, N., Shimizu, S., Fujii, Y., Umezawa, M., Takao, S. et al. (2013) Integration of a real-time tumor monitoring system into gated proton spot-scanning beam therapy: an initial phantom study using patient tumor trajectory data. Med. Phys. 40, 071729.
93) Minohara, S., Kanai, T., Endo, M., Noda, K. and Kanazawa, M. (2000) Respiratory gated irradiation system for heavy-ion radiotherapy. Int. J. Radiat. Oncol. Biol. Phys. 47, 1097–1103.
94) Zhang, Y., Trnkova, P., Toshito, T., Heijmen, B., Richter, C., Aznar, M. et al. (2023) A survey of practice patterns for real-time intrafractional motion-management in particle therapy. Phys. Imaging Radiat. Oncol. 26, 100439.
95) Shimizu, S., Matsuura, T., Umezawa, M., Hiramoto, K., Miyamoto, N., Umegaki, K. et al. (2014) Preliminary analysis for integration of spot-scanning proton beam therapy and real-time imaging and gating. Phys. Med. 30, 555–558.
96) Shimizu, S., Miyamoto, N., Matsuura, T., Fujii, Y., Umezawa, M., Umegaki, K. et al. (2014) A proton beam therapy system dedicated to spot-scanning increases accuracy with moving tumors by real-time imaging and gating and reduces equipment size. PLoS One 9, e94971.
97) Yamada, T., Miyamoto, N., Matsuura, T., Takao, S., Fujii, Y., Matsuzaki, Y. et al. (2016) Optimization and evaluation of multiple gating beam delivery in a synchrotron-based proton beam scanning system using a real-time imaging technique. Phys. Med. 32, 932–937.
98) Matsuura, T., Fujii, Y., Takao, S., Yamada, T., Matsuzaki, Y., Miyamoto, N. et al. (2016) Development and evaluation of a short-range applicator for treating superficial moving tumors with respiratory-gated spot-scanning proton therapy using real-time image guidance. Phys. Med. Biol. 61, 1515–1531.
99) Kanehira, T., Matsuura, T., Takao, S., Matsuzaki, Y., Fujii, Y., Fujii, T. et al. (2017) Impact of real-time image gating on spot scanning proton therapy for lung tumors: A simulation study. Int. J. Radiat. Oncol. Biol. Phys. 97, 173–181.
100) Yoshimura, T., Shimizu, S., Hashimoto, T., Nishioka, K., Katoh, N., Inoue, T. et al. (2020) Analysis of treatment process time for real-time-image gated-spot-scanning proton-beam therapy (RGPT) system. J. Appl. Clin. Med. Phys. 21, 38–49.
101) Kopp-Kubel, S. (1995) International Nonproprietary Names (INN) for pharmaceutical substances. Bull. World Health Organ. 73, 275–279.
102) Shirato, H., Shimizu, S., Kitamura, K. and Onimaru, R. (2007) Organ motion in image-guided radiotherapy: lessons from real-time tumor-tracking radiotherapy. Int. J. Clin. Oncol. 12, 8–16.
103) Kupelian, P., Willoughby, T., Mahadevan, A., Djemil, T., Weinstein, G., Jani, S. et al. (2007) Multi-institutional clinical experience with the Calypso System in localization and continuous, real-time monitoring of the prostate gland during external radiotherapy. Int. J. Radiat. Oncol. Biol. Phys. 67, 1088–1098.
104) Paulson, E.S., Ahunbay, E., Chen, X., Mickevicius, N.J., Chen, G.P., Schultz, C. et al. (2020) 4D-MRI driven MR-guided online adaptive radiotherapy for abdominal stereotactic body radiation therapy on a high field MR-Linac: Implementation and initial clinical experience. Clin. Transl. Radiat. Oncol. 23, 72–79.
105) Withers, H.R. and Thames, H.D. (1988) Dose fractionation and volume effects in normal tissues and tumors. Am. J. Clin. Oncol. 11, 313–329.
106) Hendry, J.H. and Thames, H.D. (1986) The tissue-rescuing unit. Br. J. Radiol. 59, 628–630.
107) Källman, P., Agren, A. and Brahme, A. (1992) Tumour and normal tissue responses to fractionated non-uniform dose delivery. Int. J. Radiat. Biol. 62, 249–262.
108) Schultheiss, T.E., Orton, C.G. and Peck, R.A. (1983) Models in radiotherapy: volume effects. Med. Phys. 10, 410–415.
109) Lyman, J.T. and Wolbarst, A.B. (1987) Optimization of radiation therapy, III: A method of assessing complication probabilities from dose-volume histograms. Int. J. Radiat. Oncol. Biol. Phys. 13, 103–109.
110) Kutcher, G.J. and Burman, C. (1989) Calculation of complication probability factors for non-uniform normal tissue irradiation: the effective volume method. Int. J. Radiat. Oncol. Biol. Phys. 16, 1623–1630.
111) Allen Li, X., Alber, M., Deasy, J.O., Jackson, A., Ken Jee, K.W., Marks, L.B. et al. (2012) The use and QA of biologically related models for treatment planning: short report of the TG-166 of the therapy physics committee of the AAPM. Med. Phys. 39, 1386–1409.
112) Fowler, J.F. (1989) The linear-quadratic formula and progress in fractionated radiotherapy. Br. J. Radiol. 62, 679–694.
113) Fowler, J.F. (2005) The radiobiology of prostate cancer including new aspects of fractionated radiotherapy. Acta Oncol. 44, 265–276.
114) Bentzen, S.M., Dörr, W., Gahbauer, R., Howell, R.W., Joiner, M.C., Jones, B. et al. (2012) Bioeffect modeling and equieffective dose concepts in radiation oncology--terminology, quantities and units. Radiother. Oncol. 105, 266–268.
115) Kwa, S.L., Theuws, J.C., Wagenaar, A., Damen, E.M., Boersma, L.J., Baas, P. et al. (1998) Evaluation of two dose-volume histogram reduction models for the prediction of radiation pneumonitis. Radiother. Oncol. 48, 61–69.
116) Graham, M.V., Purdy, J.A., Emami, B., Harms, W., Bosch, W., Lockett, M.A. et al. (1999) Clinical dose-volume histogram analysis for pneumonitis after 3D treatment for non-small cell lung cancer (NSCLC). Int. J. Radiat. Oncol. Biol. Phys. 45, 323–329.
117) Tsujino, K., Hirota, S., Endo, M., Obayashi, K., Kotani, Y., Satouchi, M. et al. (2003) Predictive value of dose-volume histogram parameters for predicting radiation pneumonitis after concurrent chemoradiation for lung cancer. Int. J. Radiat. Oncol. Biol. Phys. 55, 110–115.
118) Seppenwoolde, Y., Lebesque, J.V., de Jaeger, K., Belderbos, J.S., Boersma, L.J., Schilstra, C. et al. (2003) Comparing different NTCP models that predict the incidence of radiation pneumonitis. Normal tissue complication probability. Int. J. Radiat. Oncol. Biol. Phys. 55, 724–735.
119) Borst, G.R., Ishikawa, M., Nijkamp, J., Hauptmann, M., Shirato, H., Onimaru, R. et al. (2009) Radiation pneumonitis in patients treated for malignant pulmonary lesions with hypofractionated radiation therapy. Radiother. Oncol. 91, 307–313.
120) Dawson, L.A., Normolle, D., Balter, J.M., McGinn, C.J., Lawrence, T.S. and Ten Haken, R.K. (2002) Analysis of radiation-induced liver disease using the Lyman NTCP model. Int. J. Radiat. Oncol. Biol. Phys. 53, 810–821.
121) Prayongrat, A., Kobashi, K., Ito, Y.M., Katoh, N., Tamura, M., Dekura, Y. et al. (2019) The normal tissue complication probability model-based approach considering uncertainties for the selective use of radiation modality in primary liver cancer patients. Radiother. Oncol. 135, 100–106.
122) Philippens, M.E., Pop, L.A., Visser, A.G., Schellekens, S.A. and van der Kogel, A.J. (2004) Dose-volume effects in rat thoracolumbar spinal cord: an evaluation of NTCP models. Int. J. Radiat. Oncol. Biol. Phys. 60, 578–590.
123) Emami, B., Lyman, J., Brown, A., Coia, L., Goitein, M., Munzenrider, J.E. et al. (1991) Tolerance of normal tissue to therapeutic irradiation. Int. J. Radiat. Oncol. Biol. Phys. 21, 109–122.
124) Marks, L.B., Yorke, E.D., Jackson, A., Ten Haken, R.K., Constine, L.S., Eisbruch, A. et al. (2010) Use of normal tissue complication probability models in the clinic. Int. J. Radiat. Oncol. Biol. Phys. 76, S10–S19.
125) Grimm, J., Marks, L.B., Jackson, A., Kavanagh, B.D., Xue, J. and Yorke, E. (2021) High dose per fraction, hypofractionated treatment effects in the clinic (HyTEC): An overview. Int. J. Radiat. Oncol. Biol. Phys. 110, 1–10.
126) Hopewell, J.W., Morris, A.D. and Dixon-Brown, A. (1987) The influence of field size on the late tolerance of the rat spinal cord to single doses of X rays. Br. J. Radiol. 60, 1099–1108.
127) Bijl, H.P., van Luijk, P., Coppes, R.P., Schippers, J.M., Konings, A.W. and van der Kogel, A.J. (2002) Dose-volume effects in the rat cervical spinal cord after proton irradiation. Int. J. Radiat. Oncol. Biol. Phys. 52, 205–211.
128) Hrycushko, B., van der Kogel, A.J., Phillips, L., Folkert, M., Sayre, J.W., Vernino, S. et al. (2020) Existence of a dose-length effect in spinal nerves receiving single-session stereotactic ablative radiation therapy. Int. J. Radiat. Oncol. Biol. Phys. 106, 1010–1016.
129) Milano, M.T., Grimm, J., Niemierko, A., Soltys, S.G., Moiseenko, V., Redmond, K.J. et al. (2021) Single- and multifraction stereotactic radiosurgery dose/volume tolerances of the brain. Int. J. Radiat. Oncol. Biol. Phys. 110, 68–86.
130) Chuong, M.D., Herrera, R., Kaiser, A., Rubens, M., Romaguera, T., Alvarez, D. et al. (2022) Induction chemotherapy and ablative stereotactic magnetic resonance image-guided adaptive radiation therapy for inoperable pancreas cancer. Front. Oncol. 12, 888462.
131) Kopek, N., Holt, M.I., Hansen, A.T. and Høyer, M. (2010) Stereotactic body radiotherapy for unresectable cholangiocarcinoma. Radiother. Oncol. 94, 47–52.
132) Shirato, H., Gupta, N.K., Jordan, T.J. and Hendry, J.H. (1990) Lack of late skin necrosis in man after high-dose irradiation using small field sizes: experiences of grid therapy. Br. J. Radiol. 63, 871–874.
133) Shirato, H., Mizuta, M. and Miyasaka, K. (1995) A mathematical model of the volume effect which postulates cell migration from unirradiated tissues. Radiother. Oncol. 35, 227–231.
134) Bijl, H.P., van Luijk, P., Coppes, R.P., Schippers, J.M., Konings, A.W. and van der Kogel, A.J. (2006) Influence of adjacent low-dose fields on tolerance to high doses of protons in rat cervical spinal cord. Int. J. Radiat. Oncol. Biol. Phys. 64, 1204–1210.
135) Denham, J.W. and Hauer-Jensen, M. (2002) The radiotherapeutic injury--a complex ‘wound’. Radiother. Oncol. 63, 129–145.
136) Stone, H.B., Coleman, C.N., Anscher, M.S. and McBride, W.H. (2003) Effects of radiation on normal tissue: consequences and mechanisms. Lancet Oncol. 4, 529–536.
137) Sung, J., Sodhi, C.P., Voltaggio, L., Hou, X., Jia, H., Zhou, Q. et al. (2019) The recruitment of extra-intestinal cells to the injured mucosa promotes healing in radiation enteritis and chemical colitis in a mouse parabiosis model. Mucosal Immunol. 12, 503–517.
138) Wijerathne, H., Langston, J.C., Yang, Q., Sun, S., Miyamoto, C., Kilpatrick, L.E. et al. (2021) Mechanisms of radiation-induced endothelium damage: Emerging models and technologies. Radiother. Oncol. 158, 21–32.
139) Ong, Z.Y., Gibson, R.J., Bowen, J.M., Stringer, A.M., Darby, J.M., Logan, R.M. et al. (2010) Pro-inflammatory cytokines play a key role in the development of radiotherapy-induced gastrointestinal mucositis. Radiat. Oncol. 5, 22.
140) Shirato, H., Seppenwoolde, Y., Kitamura, K., Onimura, R. and Shimizu, S. (2004) Intrafractional tumor motion: lung and liver. Semin. Radiat. Oncol. 14, 10–18.
141) Seppenwoolde, Y., Shirato, H., Kitamura, K., Shimizu, S., van Herk, M., Lebesque, J.V. et al. (2002) Precise and real-time measurement of 3D tumor motion in lung due to breathing and heartbeat, measured during radiotherapy. Int. J. Radiat. Oncol. Biol. Phys. 53, 822–834.
142) Takao, S., Miyamoto, N., Matsuura, T., Onimaru, R., Katoh, N., Inoue, T. et al. (2016) Intrafractional baseline shift or drift of lung tumor motion during gated radiation therapy with a real-time tumor-tracking system. Int. J. Radiat. Oncol. Biol. Phys. 94, 172–180.
143) Kitamura, K., Shirato, H., Seppenwoolde, Y., Shimizu, T., Kodama, Y., Endo, H. et al. (2003) Tumor location, cirrhosis, and surgical history contribute to tumor movement in the liver, as measured during stereotactic irradiation using a real-time tumor-tracking radiotherapy system. Int. J. Radiat. Oncol. Biol. Phys. 56, 221–228.
144) Kitamura, K., Shirato, H., Seppenwoolde, Y., Onimaru, R., Oda, M., Fujita, K. et al. (2002) Three-dimensional intrafractional movement of prostate measured during real-time tumor-tracking radiotherapy in supine and prone treatment positions. Int. J. Radiat. Oncol. Biol. Phys. 53, 1117–1123.
145) Shimizu, S., Osaka, Y., Shinohara, N., Sazawa, A., Nishioka, K., Suzuki, R. et al. (2011) Use of implanted markers and interportal adjustment with real-time tracking radiotherapy system to reduce intrafraction prostate motion. Int. J. Radiat. Oncol. Biol. Phys. 81, e393–e399.
146) Hashimoto, T., Shirato, H., Kato, M., Yamazaki, K., Kurauchi, N., Morikawa, T. et al. (2005) Real-time monitoring of a digestive tract marker to reduce adverse effects of moving organs at risk (OAR) in radiotherapy for thoracic and abdominal tumors. Int. J. Radiat. Oncol. Biol. Phys. 61, 1559–1564.
147) Ahn, Y.C., Shimizu, S., Shirato, H., Hashimoto, T., Osaka, Y., Zhang, X.Q. et al. (2004) Application of real-time tumor-tracking and gated radiotherapy system for unresectable pancreatic cancer. Yonsei Med. J. 45, 584–590.
148) Katoh, N., Onimaru, R., Sakuhara, Y., Abo, D., Shimizu, S., Taguchi, H. et al. (2008) Real-time tumor-tracking radiotherapy for adrenal tumors. Radiother. Oncol. 87, 418–424.
149) Shirato, H., Shimizu, S., Taguchi, H., Takao, S., Miyamoto, N. and Matsuura, T. (2022) Tumor motion control. Springer, Berlin, Heidelberg.
150) Uchinami, Y., Kanehira, T., Fujita, Y., Miyamoto, N., Yokokawa, K., Koizumi, F. et al. (2023) Evaluation of short-term gastrointestinal motion and its impact on dosimetric parameters in stereotactic body radiation therapy for pancreatic cancer. Clin. Transl. Radiat. Oncol. 39, 100576.
151) Blomgren, H., Lax, I., Näslund, I. and Svanström, R. (1995) Stereotactic high dose fraction radiation therapy of extracranial tumors using an accelerator. Clinical experience of the first thirty-one patients. Acta Oncol. 34, 861–870.
152) Arimoto, T., Usubuchi, H., Matsuzawa, T., Yonesaka, A., Shimizu, S., Shirato, H. et al. (1998) Small volume Multiple non-coplanar Arc Radiotherapy (SMART) for tumors of the lung, head&neck and the abdominopelvic region. Elsevier, Tokyo.
153) Uematsu, M., Shioda, A., Tahara, K., Fukui, T., Yamamoto, F., Tsumatori, G. et al. (1998) Focal, high dose, and fractionated modified stereotactic radiation therapy for lung carcinoma patients: a preliminary experience. Cancer 82, 1062–1070.
154) Uematsu, M., Shioda, A., Suda, A., Fukui, T., Ozeki, Y., Hama, Y. et al. (2001) Computed tomography-guided frameless stereotactic radiotherapy for stage I non-small cell lung cancer: a 5-year experience. Int. J. Radiat. Oncol. Biol. Phys. 51, 666–670.
155) Sibley, G.S. (1998) Radiotherapy for patients with medically inoperable Stage I nonsmall cell lung carcinoma: smaller volumes and higher doses--a review. Cancer 82, 433–438.
156) Inoue, T., Katoh, N., Onimaru, R., Shimizu, S., Tsuchiya, K., Suzuki, R. et al. (2013) Stereotactic body radiotherapy using gated radiotherapy with real-time tumor-tracking for stage I non-small cell lung cancer. Radiat. Oncol. 8, 69.
157) van der Voort van Zyp, N.C., Prévost, J.B., Hoogeman, M.S., Praag, J., van der Holt, B., Levendag, P.C. et al. (2009) Stereotactic radiotherapy with real-time tumor tracking for non-small cell lung cancer: clinical outcome. Radiother. Oncol. 91, 296–300.
158) Onimaru, R., Shirato, H., Shibata, T., Hiraoka, M., Ishikura, S., Karasawa, K. et al. (2015) Phase I study of stereotactic body radiation therapy for peripheral T2N0M0 non-small cell lung cancer with PTV < 100 cc using a continual reassessment method (JCOG0702). Radiother. Oncol. 116, 276–280.
159) Onishi, H., Araki, T., Shirato, H., Nagata, Y., Hiraoka, M., Gomi, K. et al. (2004) Stereotactic hypofractionated high-dose irradiation for stage I nonsmall cell lung carcinoma: clinical outcomes in 245 subjects in a Japanese multiinstitutional study. Cancer 101, 1623–1631.
160) Ball, D., Mai, G.T., Vinod, S., Babington, S., Ruben, J., Kron, T. et al. (2019) Stereotactic ablative radiotherapy versus standard radiotherapy in stage 1 non-small-cell lung cancer (TROG 09.02 CHISEL): a phase 3, open-label, randomised controlled trial. Lancet Oncol. 20, 494–503.
161) VanderWalde, N., Jagsi, R., Dotan, E., Baumgartner, J., Browner, I.S., Burhenn, P. et al. (2016) NCCN guidelines insights: Older adult oncology, version 2.2016. J. Natl. Compr. Canc. Netw. 14, 1357–1370.
162) Onishi, H., Shirato, H., Nagata, Y., Hiraoka, M., Fujino, M., Gomi, K. et al. (2011) Stereotactic body radiotherapy (SBRT) for operable stage I non-small-cell lung cancer: can SBRT be comparable to surgery? Int. J. Radiat. Oncol. Biol. Phys. 81, 1352–1358.
163) Nagata, Y., Hiraoka, M., Shibata, T., Onishi, H., Kokubo, M., Karasawa, K. et al. (2015) Prospective trial of stereotactic body radiation therapy for both operable and inoperable T1N0M0 non-small cell lung cancer: Japan clinical oncology group study JCOG0403. Int. J. Radiat. Oncol. Biol. Phys. 93, 989–996.
164) Chang, J.Y., Mehran, R.J., Feng, L., Verma, V., Liao, Z., Welsh, J.W. et al. (2021) Stereotactic ablative radiotherapy for operable stage I non-small-cell lung cancer (revised STARS): long-term results of a single-arm, prospective trial with prespecified comparison to surgery. Lancet Oncol. 22, 1448–1457.
165) Takeda, A., Sanuki, N., Tsurugai, Y., Iwabuchi, S., Matsunaga, K., Ebinuma, H. et al. (2016) Phase 2 study of stereotactic body radiotherapy and optional transarterial chemoembolization for solitary hepatocellular carcinoma not amenable to resection and radiofrequency ablation. Cancer 122, 2041–2049.
166) Kimura, T., Takeda, A., Tsurugai, Y., Kawano, R., Doi, Y., Oku, Y. et al. (2020) A multi-institutional retrospective study of repeated stereotactic body radiation therapy for intrahepatic recurrent hepatocellular carcinoma. Int. J. Radiat. Oncol. Biol. Phys. 108, 1265–1275.
167) Taguchi, H., Sakuhara, Y., Hige, S., Kitamura, K., Osaka, Y., Abo, D. et al. (2007) Intercepting radiotherapy using a real-time tumor-tracking radiotherapy system for highly selected patients with hepatocellular carcinoma unresectable with other modalities. Int. J. Radiat. Oncol. Biol. Phys. 69, 376–380.
168) Iizuka, Y., Hiraoka, M., Kokubo, M., Sakamoto, T., Karasawa, K., Murofushi, K. et al. (2023) Dynamic tumor-tracking stereotactic body radiotherapy with real-time monitoring of liver tumors using a gimbal-mounted linac: A multi-institutional phase II study. Clin. Transl. Radiat. Oncol. 39, 100591.
169) Toramatsu, C., Katoh, N., Shimizu, S., Nihongi, H., Matsuura, T., Takao, S. et al. (2013) What is the appropriate size criterion for proton radiotherapy for hepatocellular carcinoma? A dosimetric comparison of spot-scanning proton therapy versus intensity-modulated radiation therapy. Radiat. Oncol. 8, 48.
170) Sugahara, S., Oshiro, Y., Nakayama, H., Fukuda, K., Mizumoto, M., Abei, M. et al. (2010) Proton beam therapy for large hepatocellular carcinoma. Int. J. Radiat. Oncol. Biol. Phys. 76, 460–466.
171) Mizumoto, M., Ogino, H., Okumura, T., Terashima, K., Murakami, M., Ogino, T. et al. (2023) Proton beam therapy for hepatocellular carcinoma: Multicenter prospective registry study in Japan. Int. J. Radiat. Oncol. Biol. Phys. In press.
172) Chapman, C.H., McGuinness, C., Gottschalk, A.R., Yom, S.S., Garsa, A.A., Anwar, M. et al. (2018) Influence of respiratory motion management technique on radiation pneumonitis risk with robotic stereotactic body radiation therapy. J. Appl. Clin. Med. Phys. 19, 48–57.
173) De Salles, A.A., Hariz, M., Bajada, C.L., Goetsch, S., Bergenheim, T., Selch, M. et al. (1993) Comparison between radiosurgery and stereotactic fractionated radiation for the treatment of brain metastases. Acta Neurochir. Suppl. (Wien) 58, 115–118.
174) Shirato, H., Takamura, A., Tomita, M., Suzuki, K., Nishioka, T., Isu, T. et al. (1997) Stereotactic irradiation without whole-brain irradiation for single brain metastasis. Int. J. Radiat. Oncol. Biol. Phys. 37, 385–391.
175) Aoyama, H., Shirato, H., Tago, M., Nakagawa, K., Toyoda, T., Hatano, K. et al. (2006) Stereotactic radiosurgery plus whole-brain radiation therapy vs. stereotactic radiosurgery alone for treatment of brain metastases: a randomized controlled trial. JAMA 295, 2483–2491.
176) Redmond, K.J., Gui, C., Benedict, S., Milano, M.T., Grimm, J., Vargo, J.A. et al. (2021) Tumor control probability of radiosurgery and fractionated stereotactic radiosurgery for brain metastases. Int. J. Radiat. Oncol. Biol. Phys. 110, 53–67.
177) Loo, M., Clavier, J.B., Attal Khalifa, J., Moyal, E. and Khalifa, J. (2021) Dose-response effect and dose-toxicity in stereotactic radiotherapy for brain metastases: A review. Cancers (Basel) 13, 6086.
178) Ryu, S., Pugh, S.L., Gerszten, P.C., Yin, F.F., Timmerman, R.D., Hitchcock, Y.J. et al. (2014) RTOG 0631 phase 2/3 study of image guided stereotactic radiosurgery for localized (1–3) spine metastases: phase 2 results. Pract. Radiat. Oncol. 4, 76–81.
179) Lucido, J.J., Mullikin, T.C., Abraha, F., Harmsen, W.S., Vaishnav, B.D., Brinkmann, D.H. et al. (2022) Single and multifraction spine stereotactic body radiation therapy and the risk of radiation induced myelopathy. Adv. Radiat. Oncol. 7, 101047.
180) Sahgal, A., Myrehaug, S.D., Siva, S., Masucci, G.L., Maralani, P.J., Brundage, M. et al. (2021) Stereotactic body radiotherapy versus conventional external beam radiotherapy in patients with painful spinal metastases: an open-label, multicentre, randomised, controlled, phase 2/3 trial. Lancet Oncol. 22, 1023–1033.
181) Brand, D.H., Tree, A.C., Ostler, P., van der Voet, H., Loblaw, A., Chu, W. et al. (2019) Intensity-modulated fractionated radiotherapy versus stereotactic body radiotherapy for prostate cancer (PACE-B): acute toxicity findings from an international, randomised, open-label, phase 3, non-inferiority trial. Lancet Oncol. 20, 1531–1543.
182) Tree, A.C., Ostler, P., van der Voet, H., Chu, W., Loblaw, A., Ford, D. et al. (2022) Intensity-modulated radiotherapy versus stereotactic body radiotherapy for prostate cancer (PACE-B): 2-year toxicity results from an open-label, randomised, phase 3, non-inferiority trial. Lancet Oncol. 23, 1308–1320.
183) Kishan, A.U., Ma, T.M., Lamb, J.M., Casado, M., Wilhalme, H., Low, D.A. et al. (2023) Magnetic resonance imaging-guided vs. computed tomography-guided stereotactic body radiotherapy for prostate cancer: The MIRAGE randomized clinical trial. JAMA Oncol. 9, 365–373.
184) Meier, R.M., Bloch, D.A., Cotrutz, C., Beckman, A.C., Henning, G.T., Woodhouse, S.A. et al. (2018) Multicenter trial of stereotactic body radiation therapy for low- and intermediate-risk prostate cancer: Survival and toxicity endpoints. Int. J. Radiat. Oncol. Biol. Phys. 102, 296–303.
185) Shimizu, S., Nishioka, K., Suzuki, R., Shinohara, N., Maruyama, S., Abe, T. et al. (2014) Early results of urethral dose reduction and small safety margin in intensity-modulated radiation therapy (IMRT) for localized prostate cancer using a real-time tumor-tracking radiotherapy (RTRT) system. Radiat. Oncol. 9, 118.
186) Onimaru, R., Shirato, H., Shimizu, S., Kitamura, K., Xu, B., Fukumoto, S. et al. (2003) Tolerance of organs at risk in small-volume, hypofractionated, image-guided radiotherapy for primary and metastatic lung cancers. Int. J. Radiat. Oncol. Biol. Phys. 56, 126–135.
187) Sio, T.T., Mohindra, P., Yu, N.Y., Ashman, J.B., Daniels, T.B., Merrell, K.W. et al. (2019) The search for optimal stereotactic body radiotherapy dose in inoperable, centrally located non-small-cell lung cancer continues. J. Clin. Oncol. 37, 2697–2699.
188) Timmerman, R., McGarry, R., Yiannoutsos, C., Papiez, L., Tudor, K., DeLuca, J. et al. (2006) Excessive toxicity when treating central tumors in a phase II study of stereotactic body radiation therapy for medically inoperable early-stage lung cancer. J. Clin. Oncol. 24, 4833–4839.
189) Kimura, T., Nagata, Y., Harada, H., Hayashi, S., Matsuo, Y., Takanaka, T. et al. (2017) Phase I study of stereotactic body radiation therapy for centrally located stage IA non-small cell lung cancer (JROSG10-1). Int. J. Clin. Oncol. 22, 849–856.
190) Duijm, M., van der Voort van Zyp, N.C., van de Vaart, P., Oomen-de Hoop, E., Mast, M.E., Hoogeman, M.S. et al. (2020) Predicting high-grade esophagus toxicity after treating central lung tumors with stereotactic radiation therapy using a normal tissue complication probability model. Int. J. Radiat. Oncol. Biol. Phys. 106, 73–81.
191) Jang, W.I., Bae, S.H., Kim, M.S., Han, C.J., Park, S.C., Kim, S.B. et al. (2020) A phase 2 multicenter study of stereotactic body radiotherapy for hepatocellular carcinoma: Safety and efficacy. Cancer 126, 363–372.
192) Higaki, E., Fukaya, M., Miyata, K., Kawai, R. and Abe, T. (2020) Successful two-stage operation for esophageal necrosis due to proton beam therapy followed by sorafenib in a case of large hepatocellular carcinoma. Surg. Case Rep. 6, 138.
193) Liu, J., Lee, P., McGee, H.M., Chung, V., Melstrom, L., Singh, G. et al. (2022) Advances in radiation oncology for pancreatic cancer: An updated review. Cancers (Basel) 14, 5725.
194) Hirata, T., Teshima, T., Nishiyama, K., Ogawa, K., Otani, K., Kawaguchi, Y. et al. (2015) Histopathological effects of preoperative chemoradiotherapy for pancreatic cancer: an analysis for the impact of radiation and gemcitabine doses. Radiother. Oncol. 114, 122–127.
195) Hammel, P., Huguet, F., van Laethem, J.L., Goldstein, D., Glimelius, B., Artru, P. et al. (2016) Effect of chemoradiotherapy vs. chemotherapy on survival in patients with locally advanced pancreatic cancer controlled after 4 months of gemcitabine with or without erlotinib: The LAP07 randomized clinical trial. JAMA 315, 1844–1853.
196) Hoyer, M., Roed, H., Sengelov, L., Traberg, A., Ohlhuis, L., Pedersen, J. et al. (2005) Phase-II study on stereotactic radiotherapy of locally advanced pancreatic carcinoma. Radiother. Oncol. 76, 48–53.
197) Schellenberg, D., Goodman, K.A., Lee, F., Chang, S., Kuo, T., Ford, J.M. et al. (2008) Gemcitabine chemotherapy and single-fraction stereotactic body radiotherapy for locally advanced pancreatic cancer. Int. J. Radiat. Oncol. Biol. Phys. 72, 678–686.
198) Katz, M.H.G., Shi, Q., Meyers, J., Herman, J.M., Chuong, M., Wolpin, B.M. et al. (2022) Efficacy of preoperative mFOLFIRINOX vs. mFOLFIRINOX plus hypofractionated radiotherapy for borderline resectable adenocarcinoma of the pancreas: The A021501 phase 2 randomized clinical trial. JAMA Oncol. 8, 1263–1270.
199) Bryant, J.M., Palm, R.F., Liveringhouse, C., Boyer, E., Hodul, P., Malafa, M. et al. (2022) Surgical and pathologic outcomes of pancreatic adenocarcinoma (PA) after preoperative ablative stereotactic magnetic resonance image guided adaptive radiation rherapy (A-SMART). Adv. Radiat. Oncol. 7, 101045.
200) Parikh, P.J., Lee, P., Low, D.A., Kim, J., Mittauer, K.E., Bassetti, M.F. et al. (2023) A multi-institutional phase 2 trial of ablative 5-fraction stereotactic magnetic resonance-guided on-table adaptive radiation therapy for borderline resectable and locally advanced pancreatic cancer. Int. J. Radiat. Oncol. Biol. Phys.
201) Mizuta, M., Takao, S., Date, H., Kishimoto, N., Sutherland, K.L., Onimaru, R. et al. (2012) A mathematical study to select fractionation regimen based on physical dose distribution and the linear-quadratic model. Int. J. Radiat. Oncol. Biol. Phys. 84, 829–833.
202) Brenner, D.J. (2008) The linear-quadratic model is an appropriate methodology for determining isoeffective doses at large doses per fraction. Semin. Radiat. Oncol. 18, 234–239.
203) Baujat, B., Bourhis, J., Blanchard, P., Overgaard, J., Ang, K.K., Saunders, M. et al. (2010) Hyperfractionated or accelerated radiotherapy for head and neck cancer. Cochrane Database Syst. Rev. 2010, CD002026.
204) Takeshima, T., Chamoto, K., Wakita, D., Ohkuri, T., Togashi, Y., Shirato, H. et al. (2010) Local radiation therapy inhibits tumor growth through the generation of tumor-specific CTL: its potentiation by combination with Th1 cell therapy. Cancer Res. 70, 2697–2706.
205) Chen, D.S. and Mellman, I. (2013) Oncology meets immunology: the cancer-immunity cycle. Immunity 39, 1–10.
206) Kordbacheh, T., Honeychurch, J., Blackhall, F., Faivre-Finn, C. and Illidge, T. (2018) Radiotherapy and anti-PD-1/PD-L1 combinations in lung cancer: building better translational research platforms. Ann. Oncol. 29, 301–310.
207) Rohrer Bley, C., Wolf, F., Gonçalves Jorge, P., Grilj, V., Petridis, I., Petit, B. et al. (2022) Dose- and volume-limiting late toxicity of FLASH radiotherapy in cats with squamous cell carcinoma of the nasal planum and in mini pigs. Clin. Cancer Res. 28, 3814–3823.

Non-standard abbreviation list

2DRT

two-dimensional radiotherapy

3DRT

three-dimensional radiotherapy

4DRT

four-dimensional radiotherapy

CBCT

cone-beam computed tomography

CBT

carbon beam therapy

critical segment

DTRT

dynamic tracking radiotherapy

CTV

clinical target volume

EBRT

external beam radiotherapy

EMT

electromagnetic transponder

EQD

equieffective dose

gastrointestinal

HCC

hepatocellular carcinoma

IGRT

image-guided radiotherapy

IMRT

intensity modulated radiotherapy

IPP

isotoxic planning policy

local control

MRgRT

Magnetic resonance guided radiotherapy

NSCLC

non-small cell lung cancer

NTCP

normal tissue complication probability

OAR

organ at risk

overall survival

PBT

proton beam therapy

RCT

randomized clinical trial

RGPT

real-time-image gated proton therapy

RGRT

real-time-image gated radiotherapy

RTRT

real-time tumor-tracking radiotherapy

TCP

tumor control probability

Profile

Hiroki Shirato was born in Sapporo in 1957. He graduated from Hokkaido University School of Medicine in 1981 and started training in radiation medicine, specializing in radiation oncology, under the supervision of Prof. Goro Irie and his team. He worked as a clinical fellow at the Cancer Control Agency in British Columbia, Canada, between 1986 and 1987 and as an honorable senior house officer at Christie Hospital and Holt Radium Institute, UK, between 1987 and 1988. He received his PhD degree from Hokkaido University in 1990 for biomedical research on pi-meson therapy. He studied stereotactic irradiation and organ motion, invented a real-time tumor-tracking radiotherapy in 1999, and became Full Professor of Radiation Medicine at Hokkaido University in 2006. He also invented a real-time-image–gated proton beam therapy in 2013 and became Distinguished Professor at Hokkaido University in 2016. For his accomplishments, he received the Research Front Award in 2007, Imperial Invention Prize in 2017 with his colleagues, and Japan Academy Prize in 2022. He has been a member of the International Commission on Radiation Units and Measurements (ICRU) since June 2023.

責任著者(Corresponding author)

J-STAGEへの登録はこちら（無料）