M. Rossi

Frameless Stereotactic Body Radiotherapy for Lung Cancer Using Four-Dimensional Cone Beam CT Guidance

PURPOSE To quantify the localization accuracy and intrafraction stability of lung cancer patients treated with frameless, four-dimensional (4D) cone beam computed tomography (CBCT)-guided stereotactic body radiotherapy (SBRT) and to calculate and validate planning target volume (PTV) margins to account for the residual geometric uncertainties. MATERIALS AND METHODS Sixty-five patients with small peripheral lung tumors were treated with SBRT without a body frame to 54 Gy in three fractions. For each fraction, three 4D-CBCT scans were acquired: before treatment to measure and correct the time-weighted mean tumor position, after correction to validate the correction applied, and after treatment to estimate the intrafraction stability. Patient-specific PTV margins were computed and subsequently validated using Monte Carlo error simulations. RESULTS Systematic tumor localization inaccuracies (1 SD) were 0.8, 0.8, and 0.9 mm for the left-right, craniocaudal, and anteroposterior direction, respectively. Random localization inaccuracies were 1.1, 1.1, and 1.4 mm. Baseline variations were 1.8, 2.9, and 3.0 mm (systematic) and 1.1, 1.5, and 2.0 mm (random), indicating the importance of image guidance. Intrafraction stability of the target was 1.2, 1.2, and 1.8 mm (systematic) and 1.3, 1.5, and 1.8 mm (random). Monte Carlo error simulations showed that patient-specific PTV margins (5.8-10.5 mm) were adequate for 94% of the evaluated cases (2-28 mm peak-to-peak breathing amplitude). CONCLUSIONS Frameless SBRT can be safely administered using 4D-CBCT guidance. Even with considerable breathing motion, the PTV margins can safely be kept small, allowing patients with larger tumors to benefit from the advantages of SBRT. In case bony anatomy would be used as a surrogate for tumor position, considerably larger PTV margins would be required.

COMPARISON OF DIFFERENT STRATEGIES TO USE FOUR-DIMENSIONAL COMPUTED TOMOGRAPHY IN TREATMENT PLANNING FOR LUNG CANCER PATIENTS

PURPOSE To discuss planning target volumes (PTVs) based on internal target volume (PTVITV), exhale-gated radiotherapy (PTVGating), and a new proposed midposition (PTVMidP; time-weighted mean tumor position) and compare them with the conventional free-breathing CT scan PTV (PTVConv). METHODS AND MATERIALS Respiratory motion induces systematic and random geometric uncertainties. Their contribution to the clinical target volume (CTV)-to-PTV margins differs for each PTV approach. The uncertainty margins were calculated using a dose-probability-based margin recipe (based on patient statistics). Tumor motion in four-dimensional CT scans was determined using a local rigid registration of the tumor. Geometric uncertainties for interfractional setup errors and tumor baseline variation were included. For PTVGating, the residual motion within a 30% gating (time) window was determined. The concepts were evaluated in terms of required CTV-to-PTV margin and PTV volume for 45 patients. RESULTS Over the patient group, the PTVITV was on average larger (+6%) and the PTVGating and PTVMidP smaller (-10%) than the PTVConv using an off-line (bony anatomy) setup correction protocol. With an on-line (soft tissue) protocol the differences in PTV compared with PTVConv were +33%, -4%, and 0, respectively. CONCLUSIONS The internal target volume method resulted in a significantly larger PTV than conventional CT scanning. The exhale-gated and mid-position approaches were comparable in terms of PTV. However, mid-position (or mid-ventilation) is easier to use in the clinic because it only affects the planning part of treatment and not the delivery.

Fusion of respiration-correlated PET and CT scans: correlated lung tumour motion in anatomical and functional scans

Lower lobe lung tumours in particular can move up to 2 cm in the cranio-caudal direction during the respiration cycle. This breathing motion causes image artefacts in conventional free-breathing computed tomography (CT) and positron emission tomography (PET) scanning, rendering delineation of structures for radiotherapy inaccurate. The purpose of this study was to develop a method for four-dimensional (4D) respiration-correlated (RC) acquisition of both CT and PET scans and to develop a framework to fuse these modalities. The breathing signal was acquired using a thermometer in the breathing airflow of the patient. Using this breathing signal, the acquired CT and PET data were grouped to the corresponding respiratory phases, thereby obtaining 4D CT and PET scans. Tumour motion curves were assessed in both image modalities. From these tumour motion curves, the deviation with respect to the mean tumour position was calculated for each phase. The absolute position of the centre of the tumour, relative to the bony anatomy, in the RCCT and gated PET scans was determined. This 4D acquisition and 4D fusion methodology was performed for five patients with lower lobe tumours. The peak-to-peak amplitude range in this sample group was 1-2 cm. The 3D tumour motion curve differed less than 1 mm between PET and CT for all phases. The mean difference in amplitude was less than 1 mm. The position of the centre of the tumour (relative to the bony anatomy) in the RCCT and gated PET scan was similar (difference <1 mm) when no atelectasis was present. Based on these results, we conclude that the method described in this study allows for accurate quantification of tumour motion in CT and PET scans and yields accurate respiration-correlated 4D anatomical and functional information on the tumour region.

Microscopic disease extension in three dimensions for non-small-cell lung cancer: development of a prediction model using pathology-validated positron emission tomography and computed tomography features.

PURPOSE One major uncertainty in radiotherapy planning of non-small-cell lung cancer concerns the definition of the clinical target volume (CTV), meant to cover potential microscopic disease extension (MDE) around the macroscopically visible tumor. The primary aim of this study was to establish pretreatment risk factors for the presence of MDE. The secondary aim was to establish the impact of these factors on the accuracy of positron emission tomography (PET) and computed tomography (CT) to assess the total tumor-bearing region at pathologic examination (CTV(path)). METHODS AND MATERIALS 34 patients with non-small-cell lung cancer who underwent CT and PET before lobectomy were included. Specimens were examined microscopically for MDE. The gross tumor volume (GTV) on CT and PET (GTV(CT) and GTV(PET), respectively) was compared with the GTV and the CTV at pathologic examination, tissue deformations being taken into account. Using multivariate logistic regression, image-based risk factors for the presence of MDE were identified, and a prediction model was developed based on these factors. RESULTS MDE was found in 17 of 34 patients (50%). The MDE did not exceed 26 mm in 90% of patients. In multivariate analysis, two parameters (mean CT tumor density and GTV(CT)) were significantly associated with MDE. The area under the curve of the two-parameter prediction model was 0.86. Thirteen tumors (38%, 95% CI: 24-55%) were identified as low risk for MDE, being potential candidates for reduced-intensity therapy around the GTV. In the low-risk group, the effective diameter of the GTV(CT/PET) accurately represented the CTV(path). In the high-risk group, GTV(CT/PET) underestimated the CTV(path) with, on average, 19.2 and 26.7 mm, respectively. CONCLUSIONS CT features have potential to predict the presence of MDE. Tumors identified as low risk of MDE show lower rates of disease around the GTV than do high-risk tumors. Both CT and PET accurately visualize the CTV(path) in low-risk tumors but underestimate it in high-risk tumors.

Mid-ventilation based PTV margins in Stereotactic Body Radiotherapy (SBRT): A clinical evaluation

PURPOSE Large tumor motion leads to large treatment volumes with an Internal Target Volume (ITV) based approach, whereas mid-ventilation (MidV) based Planning Target Volumes (PTV) margins typically lead to smaller treatment volumes. The purpose of this study was to evaluate the MidV approach on clinical outcome data of Stereotactic Body Radiotherapy (SBRT) in NSCLC. METHODS AND MATERIALS 297 patients with 314 peripheral tumors treated from 2006 to 2012 were retrospectively analyzed. In all patients a 4D-CT was acquired and the MidV-CT-scan was selected. Tumor amplitudes were determined in left-right (LR), cranio-caudal (CC) and anterior-posterior (AP) direction, to calculate patient specific PTV margins. RESULTS The median LR, CC and AP tumor amplitudes were 2mm (0-16 mm), 4mm (0-39 mm) and 3mm (0-18 mm), respectively, yielding a median CTV-to-PTV margin of 8mm. An ITV+5mm based PTV margin would have been bigger in 47% of the patients. After a median follow up of 22 months, local recurrence occurred in six patients (2%). Two year LC and OS were 98% and 67%, respectively. CONCLUSIONS Using the MidV approach combined with online image guidance an excellent LC of 98% was established with SBRT. This provides clinical support that incorporating respiratory motion into the PTV margin is a safe approach.

Relating acute esophagitis to radiotherapy dose using FDG-PET in concurrent chemo-radiotherapy for locally advanced non-small cell lung cancer

PURPOSE To correlate radiotherapy (RT) dose to acute esophagitis (AE) by means of FDG-PET scans acquired after concurrent chemo-radiotherapy (cCRT) for locally advanced non-small-cell lung cancer (NSCLC). MATERIALS AND METHODS Patients treated with 24 × 2.75 Gy were selected on presence of a post-RT PET (PET(post)) scan acquired within 3 months after cCRT. The value of PET(post) in relation to AE was evaluated by comparing the mean esophageal SUV of the highest 50% (mathematical left angle bracket SUV(50%) mathematical right angle bracket) between gr < 2 and gr ≥ 2AE. The local dose on the esophagus wall was correlated to the SUV and modeled using a power-law fit. The Lyman-Kutcher-Burman (LKB) model was used to predict gr ≥ 2AE. The local dose-response relation was used in the LKB model to calculate the EUD. Resulting prediction accuracy was compared to D(mean), V(35), V(55) and V(60). RESULTS Eighty-two patients were included (gr < 2 = 25, gr ≥ 2=57). The mathematical left angle bracket SUV(50%) mathematical right angle bracket ≥ was significantly higher for gr ≥ 2AE (2.2 vs. 2.6, p < 0.01). The LKB parameters (95% CI) were n = 0.130 (0.120-0.141), m = 0.25 (0.13-0.85) and TD(50) = 50.4 Gy (37.5-55.4), which resulted in improved predictability of AE compared to other predictors. CONCLUSION Esophageal uptake of FDG post-cCRT reflects AE severity. Predictability of grade ≥ 2AE was improved by using the local dose-SUV response model, with narrow confidence intervals for the optimized LKB parameters.

Local dose–effect relations for lung perfusion post stereotactic body radiotherapy

PURPOSE To model the local dose-effect relation for lung perfusion reduction in lung cancer patients treated with stereotactic body radiotherapy (SBRT). MATERIALS AND METHODS Forty-two patients having upper-lobe peripheral tumours <5 cm treated with SBRT (3×18 Gy) underwent single-photon emission computed-tomography (SPECT) scans to measure the lung perfusion 2 weeks pre-SBRT, 4-months post-SBRT, and for 8 patients 15-months post-SBRT. The relation between the calculated relative local perfusion reduction and the normalised total dose (α/β=3 Gy) at 4-months post-SBRT was modeled by 3-parameter logistic model and 2-parameter linear-maximum model. RESULTS The relation between local dose and perfusion reduction at 4-months post-SBRT showed a maximum effect of 42.6% at doses >100 Gy and was best described by the logistic model with parameters (95% CI): M=42.6% (40.7-44.6), D50=28.7 Gy (26.3-31.1) and k=2.2 (1.8-2.5). A significant increase of this maximum effect to 65.2% was found at 15-months post-SBRT. CONCLUSIONS The relation between local dose and perfusion reduction in patients treated with SBRT can be modeled by a 3-parameter logistic model. This demonstrated relationship 4-months post-SBRT approaches a plateau for doses >100 Gy, where 90% of the maximum lung-perfusion reduction is observed at NTD=78 Gy. A further perfusion reduction compared to 4-months post-SBRT was observed fifteen months post-SBRT.

Improved progression free survival for patients with diabetes and locally advanced non-small cell lung cancer (NSCLC) using metformin during concurrent chemoradiotherapy

BACKGROUND AND PURPOSE The aim was to investigate whether the use of metformin during concurrent chemoradiotherapy (cCRT) for locally advanced non-small cell lung cancer (NSCLC) improved treatment outcome. MATERIAL AND METHODS A total of 682 patients were included in this retrospective cohort study (59 metformin users, 623 control patients). All received cCRT in one of three participating radiation oncology departments in the Netherlands between January 2008 and January 2013. Primary endpoint was locoregional recurrence free survival (LRFS), secondary endpoints were overall survival (OS), progression-free survival (PFS) and distant metastasis free survival (DMFS). RESULTS No significant differences in LRFS or OS were found. Metformin use was associated with an improved DMFS (74% versus 53% at 2 years; p=0.01) and PFS (58% versus 37% at 2 years and a median PFS of 41 months versus 15 months; p=0.01). In a multivariate cox-regression analysis, the use of metformin was a statistically significant independent variable for DMFS and PFS (p=0.02 and 0.03). CONCLUSIONS Metformin use during cCRT is associated with an improved DMFS and PFS for locally advanced NSCLC patients, suggesting that metformin may be a valuable treatment addition in these patients. Evidently, our results merit to be verified in a prospective trial.

Differential motion between mediastinal lymph nodes and primary tumor in radically irradiated lung cancer patients.

PURPOSE/OBJECTIVE In patients with locally advanced lung cancer, planning target volume margins for mediastinal lymph nodes and tumor after a correction protocol based on bony anatomy registration typically range from 1 to 1.5 cm. Detailed information about lymph node motion variability and differential motion with the primary tumor, however, is lacking from large series. In this study, lymph node and tumor position variability were analyzed in detail and correlated to the main carina to evaluate possible margin reduction. METHODS AND MATERIALS Small gold fiducial markers (0.35 × 5 mm) were placed in the mediastinal lymph nodes of 51 patients with non-small cell lung cancer during routine diagnostic esophageal or bronchial endoscopic ultrasonography. Four-dimensional (4D) planning computed tomographic (CT) and daily 4D cone beam (CB) CT scans were acquired before and during radical radiation therapy (66 Gy in 24 fractions). Each CBCT was registered in 3-dimensions (bony anatomy) and 4D (tumor, marker, and carina) to the planning CT scan. Subsequently, systematic and random residual misalignments of the time-averaged lymph node and tumor position relative to the bony anatomy and carina were determined. Additionally, tumor and lymph node respiratory amplitude variability was quantified. Finally, required margins were quantified by use of a recipe for dual targets. RESULTS Relative to the bony anatomy, systematic and random errors ranged from 0.16 to 0.32 cm for the markers and from 0.15 to 0.33 cm for the tumor, but despite similar ranges there was limited correlation (0.17-0.71) owing to differential motion. A large variability in lymph node amplitude between patients was observed, with an average motion of 0.56 cm in the cranial-caudal direction. Margins could be reduced by 10% (left-right), 27% (cranial-caudal), and 10% (anteroposterior) for the lymph nodes and -2%, 15%, and 7% for the tumor if an online carina registration protocol replaced a protocol based on bony anatomy registration. CONCLUSIONS Detailed analysis revealed considerable lymph node position variability, differential motion, and respiratory motion. Planning target volume margins can be reduced up to 27% in lung cancer patients when the carina registration replaces bony anatomy registration.

The impact of microscopic disease on the tumor control probability in non-small-cell lung cancer.

PURPOSE To indicate which clinical target volume (CTV) margin (if any) is needed for an adequate treatment of non-small-cell lung cancer (NSCLC) using either 3D conformal or stereotactic radiotherapy, taking the distribution of the microscopic disease extension (MDE) into account. METHODS AND MATERIALS On the basis of the linear-quadratic biological model, a Monte-Carlo simulation was used to study the impact of MDE and setup deviations on the tumor control probability (TCP) after typical 3D conformal and stereotactic irradiation techniques. Setup deviations were properly accounted for in the planning target volume (PTV) margin. Previously measured distributions of MDE outside the macroscopic tumor in NSCLC patients were used. The dependence of the TCP on the CTV margins was quantified. RESULTS The presence of MDE had a demonstratable influence on the TCP in both the 3D conformal and the stereotactic technique when no CTV margins were employed. The impact of MDE on the TCP values was greater in the 3D conformal scenario (67% TCP with MDE; 84% TCP without MDE) than for stereotactic radiotherapy (91% TCP with MDE; 100% TCP without MDE). Accordingly, an increase of the CTV margin had the greatest impact for the 3D conformal technique. Larger setup errors, with appropriate PTV margins, lead to an increase in TCP for both techniques, showing the interdependence of CTV and PTV margins. CONCLUSIONS MDE may not always be eradicated by the beam penumbra or existing PTV margins using either 3D conformal or stereotactic radiotherapy. Nonetheless, TCP modeling indicates an overall local control rate above 90% for the stereotactic technique, while a non-zero CTV margin is recommended for better local control of MDE when using the 3D conformal technique.