Radiotherapy for mediastinal lymphoma in breath hold using surface monitoring and nasal high flow oxygen: Clinical experiences and breath hold stability

PURPOSE

In this study we describe the clinical introduction and evaluation of radiotherapy in mediastinal lymphoma in breath hold using surface monitoring combined with nasal high flow therapy (NHFT) to prolong breath hold duration.

MATERIALS AND METHODS

11 Patients with mediastinal lymphoma were evaluated. 6 Patients received NHFT, 5 patients were treated in breath hold without NHFT. Breath hold stability as measured by a surface scanning system was evaluated, as well as internal movement based on cone beam computed tomography (CBCT) before and after treatment. Based on internal movement, margins were determined. In a parallel planning study we compared free breathing plans with breath hold plans using the determined margins.

RESULTS

Average inter breath hold stability was 0.6 mm for NHFT treatments, and 0.5 mm for non-NHFT treatments (p > 0.1). Intra breath hold stability was 0.8 vs. 0.6 mm (p > 0.1) on average. Using NHFT, average breath hold duration increased from 34 s to 60 s (p < 0.01). Residual CTV motion derived from CBCTs before and after each fraction was 2.0 mm for NHFT vs 2.2 mm for non-NHFT (p > 0.1). Combined with inter-fraction motion, a uniform mediastinal margin of 5 mm appears to be sufficient. In breath hold, mean lung dose is reduced by 2.6 Gy (p < 0.001), while mean heart dose is reduced by 2.0 Gy (p < 0.001).

CONCLUSION

Treatment of mediastinal lymphoma in breath hold is feasible and safe. The addition of NHFT approximately increases breath hold durations with a factor two while stability is maintained. By reducing breathing motion, margins can be decreased to 5 mm. A considerable dose reduction in heart, lungs, esophagus, and breasts can be achieved with this method.

Automatic quality assurance of radiotherapy treatment plans using Bayesian networks: A multi-institutional study

Purpose

Artificial intelligence applications in radiation oncology have been the focus of study in the last decade. The introduction of automated and intelligent solutions for routine clinical tasks, such as treatment planning and quality assurance, has the potential to increase safety and efficiency of radiotherapy. In this work, we present a multi-institutional study across three different institutions internationally on a Bayesian network (BN)-based initial plan review assistive tool that alerts radiotherapy professionals for potential erroneous or suboptimal treatment plans.

Methods

Clinical data were collected from the oncology information systems in three institutes in Europe (Maastro clinic - 8753 patients treated between 2012 and 2020) and the United States of America (University of Vermont Medical Center [UVMMC] - 2733 patients, University of Washington [UW] - 6180 patients, treated between 2018 and 2021). We trained the BN model to detect potential errors in radiotherapy treatment plans using different combinations of institutional data and performed single-site and cross-site validation with simulated plans with embedded errors. The simulated errors consisted of three different categories: i) patient setup, ii) treatment planning and iii) prescription. We also compared the strategy of using only diagnostic parameters or all variables as evidence for the BN. We evaluated the model performance utilizing the area under the receiver-operating characteristic curve (AUC).

Results

The best network performance was observed when the BN model is trained and validated using the dataset in the same center. In particular, the testing and validation using UVMMC data has achieved an AUC of 0.92 with all parameters used as evidence. In cross-validation studies, we observed that the BN model performed better when it was trained and validated in institutes with similar technology and treatment protocols (for instance, when testing on UVMMC data, the model trained on UW data achieved an AUC of 0.84, compared with an AUC of 0.64 for the model trained on Maastro data). Also, combining training data from larger clinics (UW and Maastro clinic) and using it on smaller clinics (UVMMC) leads to satisfactory performance with an AUC of 0.85. Lastly, we found that in general the BN model performed better when all variables are considered as evidence.

Conclusion

We have developed and validated a Bayesian network model to assist initial treatment plan review using multi-institutional data with different technology and clinical practices. The model has shown good performance even when trained on data from clinics with divergent profiles, suggesting that the model is able to adapt to different data distributions.

Actual target coverage after setup verification using surgical clips compared with external skin markers in postoperative breast cancer radiation therapy

PURPOSE

After changing from offline setup verification to online setup verification using external skin markers in breast cancer patients, we noticed an increase in localized acute skin toxicity beneath the markers. Also, in vivo 3-dimensional dose measurements showed deviations between the delivered and the planned dose distributions; therefore, we investigated the accuracy of setup verification using surgical clips in the tumor bed, with a focus on target coverage of whole breast and tumor bed.

METHODS AND MATERIALS

Orthogonal kilovoltage images were acquired before every fraction in 35 breast cancer patients, deriving an online 3-dimensional setup error by matching on external skin markers. In retrospect, a rematch was performed using surgical clips. For 155 fractions (ie, 5-6 fractions/patient), a cone beam computed tomography (CT) scan was available. Analysis concerned: (1) visibility of the clips, (2) migration of the clips, (3) comparison of setup errors according to both match methods, and (4) comparison of target coverage by recalculating the dose on the online setup-corrected cone beam CT scan with the patient setup according to both match methods. External validation of the surgical clip-based online setup verification was performed in 23 patients by analyzing kilovoltage images of 100 fractions, obtained after treatment.

RESULTS

All types of surgical clips could be visualized. The clip to center-of-mass distance decreased on average by 2 mm (standard deviation, 1) over the course of treatment. Setup differences between match methods were on average <0.5 mm in all directions. The reconstructed dose distributions showed standard deviations of volumes receiving 95% or 107% of prescribed dose and mean dose of the breast and boost planning target volume were similar for the planning CT and the cone beam CTs, for both match procedures. An external validation in 23 patients showed reassuring setup errors <2 mm.

CONCLUSIONS

Online setup verification using surgical clips results in comparable setup corrections and target volume coverage as verification using skin markers. By omitting skin markers acute skin toxicity beneath the markers is prevented.

Intra-voxel heterogeneity influences the dose prescription for dose-painting with radiotherapy: a modelling study

The purpose of this study was to increase the potential of dose redistribution by incorporating estimates of oxygen heterogeneity within imaging voxels for optimal dose determination. Cellular oxygen tension (pO(2)) distributions were estimated for imaging-size-based voxels by solving oxygen diffusion-consumption equations around capillaries placed at random locations. The linear-quadratic model was used to determine cell survival in the voxels as a function of pO(2) and dose. The dose distribution across the tumour was optimized to yield minimal survival after 30 x 2 Gy fractions by redistributing the dose based on differences in oxygen levels. Eppendorf data of a series of 69 tumours were used as a surrogate of what might be expected from oxygen imaging datasets. Dose optimizations were performed both taking into account cellular heterogeneity in oxygenation within voxels and assuming a homogeneous cellular distribution of oxygen. Our simulations show that dose redistribution based on derived cellular oxygen distributions within voxels result in dose distributions that require less total dose to obtain the same degree of cell kill as dose distributions that were optimized with a model that considered voxels as homogeneous with respect to oxygen. Moderately hypoxic tumours are expected to gain most from dose redistribution. Incorporating cellular-based distributions of radiosensitivity into dose-planning algorithms theoretically improves the potential gains from dose redistribution algorithms.

Photodynamic therapy as adjuvant therapy in surgically treated pleural malignancies

Five patients with a pleural malignancy (four malignant mesotheliomas and one localized low grade carcinoid) were treated with maximal surgical resection of the tumour followed by intraoperative adjuvant photodynamic therapy (PDT). The additional photodynamic treatment was performed with light of 652 nm from a high power diode laser, and meta-tetrahydroxy phenylchlorin as the photosensitizer. The light delivery to the thoracic cavity was monitored by in situ isotropic light detectors. The position of the light delivery fibre was adjusted to achieve optimal light distribution, taking account of reflected and scattered light in this hollow cavity. There was no 30-day post-operative mortality and only one patient suffered from a major complication (diaphragmatic rupture and haematopericardium). The operation time was increased by a maximum of 1 h to illuminate the total hemithoracic surface with 10 J cm(-2) (incident and scattered light). The effect of the adjuvant PDT was monitored by examination of biopsies taken 24 h after surgery under thoracoscopic guidance. Significant damage, including necrosis, was observed in the marker lesions with remaining malignancy compared with normal tissue samples, which showed only an infiltration with PMN cells and oedema of the striated muscles cells. Of the five patients treated, four are alive with no signs of recurrent tumour with a follow-up of 9-11 months. One patient was diagnosed as having a tumour dissemination in the skin around the thoracoscopy scar and died of abdominal tumour spread. Light delivery to large surfaces for adjuvant PDT is feasible in a relatively short period of time (< 1 h). In situ dosimetry ensures optimal light distribution and allows total doses (incident plus scattered light) to be monitored at different positions within the cavity. This combination of light delivery and dosimetry is well suited for adjuvant treatment with PDT in malignant pleural tumours.

Backward and forward walking use different patterns of phase-dependent modulation of cutaneous reflexes in humans

1. The phase-dependent modulation of medium-latency (P2) (70-80 ms) responses in semitendinosus (ST), biceps femoris (BF), rectus femoris (RF), and tibialis anterior (TA) was studied with the use of low-intensity stimulation (2 times perception threshold) of the sural nerve. The shocks were given in a random order at 16 phases of the step cycle in 10 normal subjects during forward walking (FW) or backward walking (BW) on a treadmill. 2. All subjects exhibited P2 responses in all muscles studied both during BW and FW. The amplitude of the facilitatory P2 responses showed phase-dependent changes that could not have been predicted on the basis of the variations in background activity throughout the step cycle. 3. During FW, the P2 facilitatory responses in BF were large (with respect to the background activity) throughout the whole step cycle except for a short period near the end of the swing phase. In ST the responses were smaller and appeared primarily at the end of the stance phase and during the first part of the swing phase. During the second half of swing the P2 responses were basically suppressive. A modulation pattern similar to the one in ST was found in RF and TA, except that there was no reversal to suppressive responses in the swing phase in RF. Instead, a reduction in the amplitude of the facilitatory P2 responses occurred. 4. During BW, the modulation pattern recorded in the same subjects was different from the one seen during FW. Large facilitatory P2 responses were present in all muscles in middle and late swing. In the first half of stance the responses were most prominently seen in BF and RF. At the end of stance and/or at the onset of swing the facilitatory responses decreased in amplitude (BF and RF) or reversed to P2 suppressions (ST and TA). 5. We conclude that there are both facilitatory and suppressive pathways from the sural nerve to the leg muscles studied and that the balance of activity in these paths is phase dependent during both FW and BW. It is suggested that the phase-dependent modulation of P2 responses could largely rely on a central motor program. During BW the same motor program is used as during FW, but possibly running in reverse, thereby causing a shift both in the timing of the reflex reversal and in the periods of reflex suppression.