External validation of a hidden Markov model for gamma-based classification of anatomical changes in lung cancer patients using EPID dosimetry.

PURPOSE: To externally validate a hidden Markov model (HMM) for classifying gamma analysis results of in vivo electronic portal imaging device (EPID) measurements into different categories of anatomical change for lung cancer patients. Additionally, the relationship between HMM classification and deviations in dose-volume histogram (DVH) metrics was evaluated. METHODS: The HMM was developed at CHU de Québec (CHUQ), and trained on features extracted from gamma analysis maps of in vivo EPID measurements from 483 fractions (24 patients, treated with three-dimensional 3D-CRT or intensity modulated radiotherapy), using the EPID measurement of the first treatment fraction as reference. The model inputs were the average gamma value, standard deviation, and average value of the highest 1% of gamma values, all averaged over all beams in a fraction. The HMM classified each fraction into one of three categories: no anatomical change (Category 1), some anatomical change (no clinical action needed, Category 2) and severe anatomical change (clinical action needed, Category 3). The external validation dataset consisted of EPID measurements from 263 fractions of 30 patients treated at Maastro with volumetric modulated arc therapy (VMAT) or hybrid plans (containing both static beams and VMAT arcs). Gamma analysis features were extracted in the same way as in the CHUQ dataset, by using the EPID measurement of the first fraction as reference (γQ), and additionally by using an EPID dose prediction as reference (γM). For Maastro patients, cone beam computed tomography (CBCT) scans and image-guided radiotherapy (IGRT) classification of these images were available for each fraction. Contours were propagated from the planning CT to the CBCTs, and the dose was recalculated using a Monte Carlo dose engine. Dose-volume histogram metrics for targets and organs-at-risk (OARs: lungs, heart, mediastinum, spinal cord, brachial plexus) were extracted for each fraction, and compared to the planned dose. HMM classification of the external validation set was compared to threshold classification based on the average gamma value alone (a surrogate for clinical classification at CHUQ), IGRT classification as performed at Maastro, and differences in DVH metrics extracted from 3D dose recalculations on the CBCTs. RESULTS: The HMM achieved 65.4%/65.0% accuracy for γQ and γM, respectively, compared to average gamma threshold classification. When comparing HMM classification with IGRT classification, the overall accuracy was 29.7% for γQ and 23.2% for γM. Hence, HMM classification and IGRT classification of anatomical changes did not correspond. However, there is a trend towards higher deviations in DVH metrics with classification into higher categories by the HMM for large OARs (lungs, heart, mediastinum), but not for the targets and small OARs (spinal cord, brachial plexus). CONCLUSION: The external validation shows that transferring the HMM for anatomical change classification to a different center is challenging, but can still be valuable. The HMM trained at CHUQ cannot be used directly to classify anatomical changes in the Maastro data. However, it may be possible to use the model in a different capacity, as an indicator for changes in the 3D dose based on two-dimensional EPID measurements.

The association of intraprostatic calcifications and dosimetry parameters with biochemical control after permanent prostate implant.

PURPOSE: The objective of this study was to evaluate the impact of intraprostatic calcifications (IC) on long-term tumor control in patients treated with permanent implant prostate brachytherapy (PIPB). MATERIALS AND METHODS: Data from 609 I-125 patients treated with PIPB were retrospectively reviewed. The presence of IC was determined by reviewing postimplant CT images. Doses delivered were determined using the Monte Carlo (model-based) calculations and the TG43 approach. Biochemical relapses at 7 and 10 years were determined according to Phoenix definition. Long-term biochemical relapse-free survival (bRFS) was determined using Kaplan-Meier estimates with log rank test. Cox proportional hazard models were used for analysis of predictor factors of biochemical recurrence. RESULTS: IC were observed for 11.1% of patients. Clinical stage, PSA, Gleason score, D'Amico risk group, and ADT use were comparable between IC and no IC groups. The 7- and 10-year bRFS for the entire cohort were 94.1% and 90.6%, respectively. The bRFS at 7 years was 90.5% (with IC) vs. 94.5% (without IC) (p = 0.198); the corresponding values at 10 years were 78.8% vs. 91.8% (p = 0.046). On Cox model, only prostatic calcifications were a significant risk factor for biochemical relapse (HR: 2.30, IC 95%: 1.05-5.00, p = 0.037; and HR: 3.94; IC 95%: 1.00-15.38; p = 0.049 for univariate and multivariate analysis, respectively). CONCLUSION: The presence of IC in patients treated with PIPB decreases V100 and D90 for postimplant Monte Carlo dosimetry (compared with TG43); correspondingly, IC are associated with a lower 10-y bRFS. Model-based dose calculations are critical to evaluate potential cold spots due to calcifications.

Radiation Therapy-Induced Dysfunction in Cardiovascular Implantable Electronic Devices.

PURPOSE: The prevalence of patients with cardiovascular implantable electronic devices (CIEDs) who receive radiation treatment for cancer is increasing. External beam radiation therapy (RT) can affect the electronic components. This study aimed to evaluate the incidence and predictors of new onset CIED dysfunction in patients treated with RT. METHODS AND MATERIALS: We retrospectively analyzed data from 230 patients with CIEDs who received radiation treatment at the Radiation Oncology Center of CHU de Québec - Université Laval between February 2007 and November 2013. The reviewed data included baseline characteristics, CIEDs, and RT treatment specifications. Patients with CIEDs were analyzed before, during, and at the end of radiation treatment. High- and low-energy photon or electron beam radiation from linear accelerators, orthovoltage machines, and high-dose rate brachytherapy delivery were used. Abnormal events could be one of the following: total or partial deprogramming of the CIED parameters, onset of new symptoms, or new arrhythmia. RESULTS: This study is based on one of the largest cohorts. A total of 18 events in 16 patients (7.8%) were recorded. Of the 18 events, 16 had at least part of the radiation treatment delivered with photo neutrons producing high-energy RT (neutron producing RT). Only 2 abnormal events occurred during non-neutron producing RT. Both the prescription dose and the dose estimated at the location of the pacemaker were correlated with the probability of an abnormal event (P = .0006 and P = .003, respectively). Among the 16 patients, clinical symptoms were noted in only 1 patient (6.3%). CONCLUSIONS: CIED malfunctions are relatively uncommon and do not seem to be life threatening. We recommend limiting the dose at the CIED and avoid neutron-producing RT to reduce the risk of CIED malfunction.

Thermoplastic wrapped lead sheet to reduce cardiac device cumulative dose.

The objective of this project is to evaluate the percentage dose reduction in cardiac implantable electronic devices (CIEDs) using a thermoplastic wrapped lead sheet. The dose to CIED is evaluated in various situations with and without a lead shield. The efficiency of this type of shielding is supported by measurements made with a commercial plastic scintillation detector (PSD). Percentage depth dose (PDD) curve and lateral dose measurements (LDMs) were made with and without shielding for photon and electron beams. Photon LDMs were made at a depth of 0.5 cm. PSD measurements were compared with dose calculation from the treatment planning system (TPS). The benefit of shielding is greater at 23 MV than at 6 MV, with an average reduction of 71% and 59% of dose, respectively, for out-of-field distance range between 3 and 15 cm. Measurement of posterior beams shows there is no significant increase in skin dose due to backscatter from the lead sheet even when the field intercepts it. Large deviations between TPS calculation and measurements have been observed. The use of lead shielding with an anterior field is advised and provides an easy way to decrease the cumulative dose to CIEDs. Interception of shielding by an electron beam would increase significantly the cumulative dose to CIED for high energies or decrease the quality of the treatment. For a posterior out-of-field, shielding does not have a significant impact on CIED dose.

Establishing action threshold for change in patient anatomy using EPID gamma analysis and PTV coverage for head and neck radiotherapy treatment.

PURPOSE: To present a new adaptive radiotherapy (ART) method based on relative gamma analysis and patient classification for the identification of anatomical changes that induce a sufficient dosimetric impact to affect the treatment delivery and require complete replanning. METHODS: This retrospective study includes 55 patients treated for a head and neck cancer with IMRT, VMAT, or 3D conformal RT. Electronic Portal Imaging Device images for all treatment fields were acquired daily at every fraction. CBCTs were collected at least once a week. Gamma analysis was performed using the first fraction of the treatment as a reference once validated that it was delivered without error. Gamma analysis parameters (<γ>, standard deviation and the Top 1% γ) were used to define categories using statistic from a k-means clustering analysis. From these categories an action threshold was defined and correlated with dosimetric changes. For 23 of 55 patients, the V100% for PTV was computed for both, the planning CT and original contours deformed onto CBCT acquired at the last fraction. These values were then compared with 2D image relative γ-analysis of EPID images. Sensitivity and specificity of the method for the detection of dosimetric changes were computed. RESULTS: Three categories indicating an increasing level of change with the planned treatment were identified. A threshold was established for which patients were at risk of deviation at <γ> = 0.42. From 23 recomputing plans, it has been confirmed that patients with a strong dosimetric impact were above this threshold, with a specificity of 0.80 and a sensitivity of 0.84. CONCLUSIONS: The specificity and the sensitivity value confirmed the performance of the method to detect anatomical changes. The γ-analysis threshold correlated well with morphological changes that have a relevant dosimetric impact. Analysis of daily EPID images provides a method to identify patients at risk of deviation from their planned treatment and can support an early replanning decision.

Classification of changes occurring in lung patient during radiotherapy using relative γ analysis and hidden Markov models.

PURPOSE: To present a new automated patient classification method based on relative gamma analysis and hidden Markov models (HMM) to identify patients undergoing important anatomical changes during radiation therapy. METHODS: Daily EPID images of every treatment field were acquired for 52 patients treated for lung cancer. In addition, CBCT were acquired on a regular basis. Gamma analysis was performed relative to the first fraction given that no significant anatomical change was observed on the CBCT of the first fraction compared to the planning CT. Several parameters were extracted from the gamma analysis (e.g., average gamma value, standard deviation, percent above 1). These parameters formed patient-specific time series. Data from the first 24 patients were used as a training set for the HMM. The trained HMM was then applied to the remaining 28 patients and compared to manual clinical evaluation and fixed thresholds. RESULTS: A three-category system was used for patient classification ranging from minor deviations (category 1) to severe deviations (category 3) from the treatment plan. Patient classified using the HMM lead to the same result as the classification made by a human expert 83% of the time. The HMM overestimate the category 10% of the time and underestimate 7% of the time. Both methods never disagree by more than one category. In addition, the information provided by the HMM is richer than the simple threshold-based approach. HMM provides information on the likelihood that a patient will improve or deteriorate as well as the expected time the patient will remain in that state. CONCLUSION: We showed a method to classify patients during the course of radiotherapy based on relative changes in EPID images and a hidden Markov model. Information obtained through this automated classification can complement the clinical information collected during treatment and help identify patients in need of a plan adaptation.

Radiotherapy-Induced Cardiac Implantable Electronic Device Dysfunction in Patients With Cancer.

Radiotherapy can affect the electronic components of a cardiac implantable electronic device (CIED) resulting in malfunction and/or damage. We sought to assess the incidence, predictors, and clinical impact of CIED dysfunction (CIED-D) after radiotherapy for cancer treatment. Clinical characteristics, cancer, different types of CIEDs, and radiation dose were evaluated. The investigation identified 230 patients, mean age 78 ± 8 years and 70% were men. A total of 199 patients had pacemakers (59% dual chamber), 21 (9%) cardioverter-defibrillators, and 10 (4%) resynchronizators or defibrillators. The left pectoral (n = 192, 83%) was the most common CIED location. Sixteen patients (7%) experienced 18 events of CIED-D after radiotherapy. Reset to backup pacing mode was the most common encountered dysfunction, and only 1 (6%) patient of those with CIED-D experienced symptoms of atrioventricular dyssynchrony. Those who had CIED-D tended to have a shorter device age at the time of radiotherapy compared to those who did not (2.5 ± 1.5 vs 3.8 ± 3.4 years, p = 0.09). The total dose prescribed to the tumor was significantly greater among those who had CIED-D (66 ± 30 vs 42 ± 23 Gy, p <0.0001). Multivariate logistic regression analysis identified the total dose prescribed to the tumor as the only independent predictor for CIED-D (odds ratio 1.19 for each increase in 5 Gy, 95% confidence interval 1.08 to 1.31, p = 0.0005). In conclusion, in this large population of patients with CIEDs undergoing radiotherapy for cancer treatment, the occurrence of newly diagnosed CIED-D was 7%, and the reset to backup pacing mode was the most common encountered dysfunction. The total dose prescribed to the tumor was a predictor of CIED-D. Importantly, although the unpredictability of CIEDs under radiotherapy is still an issue, none of our patients experienced significant symptoms, life-threatening arrhythmias, or conduction disorders.

Technical Note: Out-of-field dose measurement at near surface with plastic scintillator detector.

Out-of-field dose depends on multiple factors, making peripheral dosimetry com-plex. Only a few dosimeters have the required features for measuring peripheral dose. Plastic scintillator dosimeters (PSDs) offer numerous dosimetric advantages as required for out-of-field dosimetry. The purpose of this study is to determine the potential of using PSD as a surface peripheral dosimeter. Measurements were performed with a parallel-plate ion chamber, a small volume ion chamber, and with a PSD. Lateral-dose measurements (LDM) at 0.5 cm depth and depth-dose curve (PDD) were made and compared to the dose calculation provided by a treatment planning system (TPS). This study shows that a PSD can measure a dose as low as 0.51 ± 0.17 cGy for photon beam and 0.58 ± 0.20 cGy for electron beam with a difference of 0.2 and 0.1 cGy compared to a parallel-plate ion chamber. This study demonstrates the potential of using PSD as an out-of-field dosimeter since measure-ments with PSD avoid averaging over a too-large depth, at 1 mm diameter, and can make precise measurement at very low dose. Also, electronic equilibrium is easier to reach with PSD due to its small sensitive volume and its water equivalence.

Response to Re: Estimating and reducing dose received by cardiac devices for patients undergoing radiotherapy.

In reply to Dimitris N. Mihailidis regarding our manuscript entitled "Estimating and reducing dose received by cardiac devices for patients undergoing radiotherapy".

Comparison of dose and catheter optimization algorithms in prostate high-dose-rate brachytherapy.

PURPOSE: The purpose of this work was to compare the hybrid inverse treatment planning optimization (HIPO), inverse dose-volume histogram-based optimization (DVHO), and fast simulated annealing stochastic algorithm (IPSA). The catheter optimization algorithm HIPO was also compared with the Centroidal Voronoi Tessellation (CVT) algorithm. METHODS AND MATERIALS: In this study, eight high-dose-rate prostate cases were randomly selected from an anonymized bank of patients. Oncentra Prostate v4.1 was used to run DVHO and the HIPO catheter optimization (HIPO_cat), whereas Oncentra Brachy v4.3 was used for the remaining. For fixed catheter configurations, DVHO plans were compared with IPSA and HIPO. For catheter positions optimization, CVT and HIPO_cat algorithms were compared with standard clinical template plans. CVT catheters were further restrained to the template grid (CVT_grid) and compared with HIPO_cat. RESULTS: For dose optimization, IPSA and HIPO were not different from each other. The urethra D10 and the computation time were found significantly better with IPSA and HIPO compared with DVHO (p < 0.0001). All other dosimetric indices were not statistically different from each others (p > 0.05). For catheter placement, CVT plans were better, whereas HIPO_cat plans were significantly worse (p < 0.05) than standard clinical plans. CVT_grid plans were similar to clinical plans and fulfilling American Brachytherapy Society guidelines down to 12 catheters, whereas HIPO_cat plans do not for all catheter numbers. The CVT algorithm run time was significantly faster than HIPO_cat (p < 0.0001). CONCLUSIONS: Dose optimization engines IPSA, DVHO, and HIPO give similar dosimetric results. The CVT approach was found to be better than HIPO_cat and was able to reduce the number of catheters significantly.

Estimating and reducing dose received by cardiac devices for patients undergoing radiotherapy.

The objectives of this project are to quantify the dose reduction effect provided by a lead shield for patients with cardiac implantable electronic devices (CIED) during a clinically realistic radiation treatment on phantom and to provide a simple model of dose estimation to predict dose received by CIED in a wide range of situations. The shield used in this project is composed of a lead sheet wrapped in thermoplastic. Dose measurements were made with a plastic scintillation detector (PSD). The phantom was treated with ten different plans. Three of these cases were treated with intensity-modulated radiation therapy (IMRT) and the others received standard 3D conformal radiation therapy (3D CRT). Lateral dose measurement for photon fields was made to establish a dose prediction model. On average, the use of the lead shield reduced the dose to CIEDs by 19% ± 13%. Dose reduction was most important for breast cases, with a mean reduction of 31% ± 15%. In three cases, the total dose reduction was more than 25 cGy over the complete treatment. For the three IMRT cases, the mean dose reduction was 11% ± 9%. On average, the difference between the TPS prediction and the measurement was 71%, while it was only 14% for the dose prediction model. It was demonstrated that a lead shield can be efficiently used for reducing doses to CIED with a wide range of clinical plans. In patients treated with IMRT modality treatment, the shielding should be used only for those with more than two anterior fields over seven fields. In the case of 3D CRT patients, the shielding should be used for those with a dose on the CIED higher than 50 cGy and with a reduction of dose higher than 10 cGy. The dose prediction model developed in this study can be an easy way to have a better estimation of the out-of-field dose than the TPS.

Characterization of lung tumors motion baseline using cone-beam computed tomography.

PURPOSE: To characterize the interfractional variability in lung tumor volume, position, and tumor boundaries. METHODS: Cone-beam computed tomography (CBCT) scans were acquired weekly during the course of treatment for 34 lung cancer patients (1-20 scans) with large tumors. Spatial registration based on bones was performed between contoured planning CT and CBCT. Gross tumor volume (GTV) on each CBCT was then contoured. Tumor volume, centroid, and boundaries variability were quantified. A commercial deformable registration software was tested and results were compared to manual contours. RESULTS: Mean volume reduction was 41 ± 32% (p < 0.001) after an average time of 51 days. Tumor centroid drifts were 0.03, 0.14, and -0.13 cm in right-left (RL), anterior-posterior (AP), and superior-inferior (SI) directions with standard deviations of 0.55, 0.50, and 0.51 cm. GTV boundaries displacements were -0.27, -0.14, and -0.16 cm with standard deviations of 0.64, 0.57, and 0.59 cm in RL, AP, and SI directions. Relative error between deformed and manual contours with the commercial deformable registration software rose up exponentially with the GTV decrease. CONCLUSIONS: GTV size changes for large lung tumors are similar to those for standard tumors. Magnitude absolute values of displacement vector for centroid and boundaries shifts show that there is not a preferred direction for the drifts but shrinkage.

Dose escalation to the dominant intraprostatic lesion defined by sextant biopsy in a permanent prostate I-125 implant: a prospective comparative toxicity analysis.

PURPOSE: Using real-time intraoperative inverse-planned permanent seed prostate implant (RTIOP/PSI), multiple core biopsy maps, and three-dimensional ultrasound guidance, we planned a boost volume (BV) within the prostate to which hyperdosage was delivered selectively. The aim of this study was to investigate the potential negative effects of such a procedure. METHODS AND MATERIALS: Patients treated with RTIOP/PSI for localized prostate cancer with topographic biopsy results received an intraprostatic boost (boost group [BG]). They were compared with patients treated with a standard plan (reference group [RG]). Plans were generated using a simulated annealing inverse planning algorithm. Prospectively recorded urinary, rectal, and sexual toxicities and dosimetric parameters were compared between groups. RESULTS: The study included 120 patients treated with boost technique who were compared with 70 patients treated with a standard plan. Boost technique did not significantly change the number of seeds (55.1/RG vs. 53.6/BG). The intraoperative prostate V150 was slightly higher in BG (75.2/RG vs. 77.2/BG, p = 0.039). Urethra V100, urethra D90, and rectal D50 were significantly lower in the BG. No significant differences were seen in acute or late urinary, rectal, or sexual toxicities. CONCLUSIONS: Because there were no differences between the groups in acute and late toxicities, we believe that BV can be planned and delivered to the dominant intraprostatic lesion without increasing toxicity. It is too soon to say whether a boost technique will ultimately increase local control.

Bypassing the learning curve in permanent seed implants using state-of-the-art technology.

PURPOSE: The aim of this study was to demonstrate, based on clinical postplan dose distributions, that technology can be used efficiently to eliminate the learning curve associated with permanent seed implant planning and delivery. METHODS AND MATERIALS: Dose distributions evaluated 30 days after the implant of the initial 22 consecutive patients treated with permanent seed implants at two institutions were studied. Institution 1 (I1) consisted of a new team, whereas institution 2 (I2) had performed more than 740 preplanned implantations over a 9-year period before the study. Both teams had adopted similar integrated systems based on three-dimensional (3D) transrectal ultrasonography, intraoperative dosimetry, and an automated seed delivery and needle retraction system (FIRST, Nucletron). Procedure time and dose volume histogram parameters such as D90, V100, V150, V200, and others were collected in the operating room and at 30 days postplan. RESULTS: The average target coverage from the intraoperative plan (V100) was 99.4% for I1 and 99.9% for I2. D90, V150, and V200 were 191.4 Gy (196.3 Gy), 75.3% (73.0%), and 37.5% (34.1%) for I1 (I2) respectively. None of these parameters shows a significant difference between institutions. The postplan D90 was 151.2 Gy for I1 and 167.3 Gy for I2, well above the 140 Gy from the Stock et al. analysis, taking into account differences at planning, results in a p value of 0.0676. The procedure time required on average 174.4 min for I1 and 89 min for I2. The time was found to decrease with the increasing number of patients. CONCLUSION: State-of-the-art technology enables a new brachytherapy team to obtain excellent postplan dose distributions, similar to those achieved by an experienced team with proven long-term clinical results. The cost for bypassing the usual dosimetry learning curve is time, with increasing team experience resulting in shorter treatment times.

Evaluation of an automatic needle-loading system.

The purpose of this paper is to evaluate the dosimetric capabilities and the radiation protection (RP) performance of a new automatic needle-loading system for permanent prostate implants, the Isoloader (Mentor Corp.). The unit has been used in more than 100 clinical cases at our institution. The Isoloader is a computerized workstation that allows automated seed testing by a solid-state CdZnTe radiation detector and loading in surgical needles. The seeds are received in a shielded and ready-to-use cartridge. Radiation protection measurements were done on a cartridge filled with 67 (125)I seeds and during dosimetric seed verification and needle loading. The reproducibility of the detector was tested and its accuracy was determined by comparison to specified activities of six calibration seeds and to their measurements in a calibrated well-chamber (WC). Finally, the times required to complete dosimetric verification and needle loading were evaluated. The cartridge was found to be adequately shielded, since no significant amount of radiation was detected around it. Radiation during seed assay was found to be worst at the cartridge's bottom, where it has a value of 15.2 microSv/h (1.4 microSv/h at 10 cm). For the needle-loading task, measurements were performed with a typical needle (three seeds) at the shielded needle holder surface yielding 307.2 microSv/h (8.3 microSv/h at 20 cm). Seed dosimetric verification takes an average of 15 s/seed, while it takes a mean time of 50 s/needle to complete the loading task. Measurements of the six seed activities were within 0.65% of the ordered activities and 1.9% higher on average than those from the WC (min = 0.7%; max = 3.5%). The reproducibility of the measurements of the CdZnTe detector was excellent, with an average of 0.01% of deviation from a reference measurement (N = 120; = 1.9%). We therefore conclude that the Isoloader is a safe, fast, and effective needle-loading system.