Impact of a commercially available model-based dose calculation algorithm on treatment planning of high-dose-rate brachytherapy in patients with cervical cancer

The effect of vaginal cylinder inhomogeneity on the HDR brachytherapy dose calculations using Monte Carlo simulations

PURPOSE

To analytically assess the heterogeneity effect of vaginal cylinders (VC) made of high-density plastics on dose calculations, considering the prescription point (surface or 5 mm beyond the surface), and benchmark the accuracy of a commercial model-based dose calculation (MBDC) algorithm using Monte Carlo (MC) simulations.

METHODS AND MATERIALS

The GEANT4 MC code was used to simulate a commercial 192 Ir HDR source and VC, with diameters ranging from 20 to 35 mm, inside a virtual water phantom. Standard plans were generated from a commercial treatment planning system [TPS-BrachyVision ACUROS (BV)] optimized for a treatment length of 5 cm through two dose calculation approaches: (1) assuming all the environment as water (i.e., Dw,w-MC & Dw,w-TG43 ) and (2) accounting for the heterogeneity of VC applicators (i.e., Dw,w-App-MC & Dw,w-App-MBDC ). The compared isodose lines, and dose & energy difference maps were extracted for analysis. In addition, the dose difference on the peripheral surface, along the applicator and at middle of treatment length, as well as apical tip was evaluated.

RESULTS

The Dw,w-App-MC results indicated that the VC heterogeneity can cause a dose reduction of (up to) % 6.8 on average (for all sizes) on the peripheral surface, translating to 1 mm shrinkage of the isodose lines compared to Dw,w-MC . In addition, the results denoted that BV overestimates the dose on the peripheral surface and apical tip of about 3.7% and 17.9%, respectively, (i.e., Dw,w-App-MBDC vs Dw,w-App-MC ) when prescribing to the surface. However, the difference between the two were negligible at the prescription point when prescribing to 5 mm beyond the surface.

CONCLUSION

The VCs' heterogeneity could cause dose reduction when prescribing dose to the surface of the applicator, and hence increases the level of uncertainty. Thus, reviewing the TG43 results, in addition to ACUROS, becomes prudent, when evaluating the surface coverage at the apex.

Automated treatment planning framework for brachytherapy of cervical cancer using 3D dose predictions

Objective. To lay the foundation for automated knowledge-based brachytherapy treatment planning using 3D dose estimations, we describe an optimization framework to convert brachytherapy dose distributions directly into dwell times (DTs).Approach. A dose rate kernelḋ(r,θ,φ)was produced by exporting 3D dose for one dwell position from the treatment planning system and normalizing by DT. By translating and rotating this kernel to each dwell position, scaling by DT and summing over all dwell positions, dose was computed (Dcalc). We used a Python-coded COBYLA optimizer to iteratively determine the DTs that minimize the mean squared error betweenDcalcand reference doseDref, computed using voxels withDref80%-120% of prescription. As validation of the optimization, we showed that the optimizer replicates clinical plans whenDref= clinical dose in 40 patients treated with tandem-and-ovoid (T&O) or tandem-and-ring (T&R) and 0-3 needles. Then we demonstrated automated planning in 10 T&O usingDref= dose predicted from a convolutional neural network developed in past work. Validation and automated plans were compared to clinical plans using mean absolute differences (MAD=1N∑n=1Nabsxn-xn') over all voxels (xn= Dose,N= #voxels) and DTs (xn= DT,N= #dwell positions), mean differences (MD) in organD2ccand high-risk CTV D90 over all patients (where positive indicates higher clinical dose), and mean Dice similarity coefficients (DSC) for 100% isodose contours.Main results. Validation plans agreed well with clinical plans (MADdose= 1.1%, MADDT= 4 s or 0.8% of total plan time,D2ccMD = -0.2% to 0.2% and D90 MD = -0.6%, DSC = 0.99). For automated plans, MADdose= 6.5% and MADDT= 10.3 s (2.1%). The slightly higher clinical metrics in automated plans (D2ccMD = -3.8% to 1.3% and D90 MD = -5.1%) were due to higher neural network dose predictions. The overall shape of the automated dose distributions were similar to clinical doses (DSC = 0.91).Significance. Automated planning with 3D dose predictions could provide significant time savings and standardize treatment planning across practitioners, regardless of experience.

Review on Treatment Planning Systems for Cervix Brachytherapy (Interventional Radiotherapy): Some Desirable and Convenient Practical Aspects to Be Implemented from Radiation Oncologist and Medical Physics Perspectives

Intracavitary brachytherapy (BT, Interventional Radiotherapy, IRT), plays an essential role in the curative intent of locally advanced cervical cancer, for which the conventional approach involves external beam radiotherapy with concurrent chemotherapy followed by BT. This work aims to review the different methodologies used by commercially available treatment planning systems (TPSs) in exclusive magnetic resonance imaging-based (MRI) cervix BT with interstitial component treatments. Practical aspects and improvements to be implemented into the TPSs are discussed. This review is based on the clinical expertise of a group of radiation oncologists and medical physicists and on interactive demos provided by the software manufacturers. The TPS versions considered include all the new tools currently in development for future commercial releases. The specialists from the supplier companies were asked to propose solutions to some of the challenges often encountered in a clinical environment through a questionnaire. The results include not only such answers but also comments by the authors that, in their opinion, could help solve the challenges covered in these questions. This study summarizes the possibilities offered nowadays by commercial TPSs, highlighting the absence of some useful tools that would notably improve the planning of MR-based interstitial component cervix brachytherapy.

Comparison of the Dosimetric Influence of Applicator Displacement on 2D and 3D Brachytherapy for Cervical Cancer Treatment

To compare the dosimetric influence of applicator displacement on two-dimensional brachytherapy (2D-BT) and three-dimensional brachytherapy (3D-BT) for cervical cancer. Nineteen patients who received computed tomography-guided tandem-and-ovoid (T&O) brachytherapy were retrospectively selected. Both 2D (point-based) and 3D (volume-based) plans with and without virtual applicator displacement in the 3 axes were created for each patient. Dose changes at point A, D₉₀ of the high-risk clinical target volume (HR-CTV) and intermediate-risk CTV (IR-CTV), and the D_0.1cc, D_1cc, D_2cc, and D_5cc of organs-at-risk (OARs) caused by applicator displacement were evaluated. Both 2D-BT and 3D-BT plans were sensitive to T&O applicator displacement. The D₉₀ of the CTV and the dose at point A were very sensitive to applicator displacement in the right–left direction (X-axis). An applicator shift of >2 mm in the X-axis resulted in a change of >5% in the dose at point A and D₉₀ of HR-CTV and IR-CTV. In addition, the doses to the OARs were mostly affected by applicator displacement in the anterior–posterior direction (Z-axis). A displacement of <1.5 mm in the Z-axis was required to avoid a dose change of >10% for OARs. For both 2D-BT and 3D-BT plans, T&O displacement greater than  ± 2 mm in the X-axis or T&O applicator displacement  ± 1.5 mm in the Z-axis resulted in significant dose changes to the tumor and OARs. In comparison with 3D-BT plans, 2D-BT plans delivered a higher dose to the tumor, and the OARs received more undesirable doses when applicator displacement occurred. The influence of applicator displacement on the doses to the tumor and OARs differed between 2D-BT and 3D-BT. Physicians should take individual patient differences into account when selecting a brachytherapy plan to mitigate the influence of applicator displacement.

A review of dosimetric impact of implementation of model-based dose calculation algorithms (MBDCAs) for HDR brachytherapy

To obtain dose distributions more physically representative to the patient anatomy in brachytherapy, calculation algorithms that can account for heterogeneity are required. The current standard AAPM Task Group No 43 (TG-43) dose calculation formalism has some clinically relevant dosimetric limitations. Lack of tissue heterogeneity and scattered dose corrections are the major weaknesses of the TG-43 formalism and could lead to systematic dose errors in target volumes and organs at risk. Over the last decade, model-based dose calculation algorithms (MBDCAs) have been clinically offered as complementary algorithms beyond the TG43 formalism for high dose rate (HDR) brachytherapy treatment planning. These algorithms provide enhanced dose calculation accuracy by using the information in the patient's computed tomography images, which allows modeling the patient's geometry, material compositions, and the treatment applicator. Several researchers have investigated the implementation of MBDCAs in HDR brachytherapy for dose optimization, but moving toward using them as primary algorithms for dose calculations is still lagging. Therefore, an overview of up-to-date research is needed to familiarize clinicians with the current status of the MBDCAs for different cancers in HDR brachytherapy. In this paper, we review the MBDCAs for HDR brachytherapy from a dosimetric perspective. Treatment sites covered include breast, gynecological, lung, head and neck, esophagus, liver, prostate, and skin cancers. Moreover, we discuss the current status of implementation of MBDCAs and the challenges.

Validation of the collapsed cone algorithm for HDR liver brachytherapy against Monte Carlo simulations

PURPOSE

To validate the collapsed cone (CC) algorithm against Monte Carlo (MC) simulations for model-based dose calculations in high-dose-rate (HDR) liver brachytherapy.

METHODS AND MATERIALS

Doses for liver brachytherapy treatment plans of 10 cases were retrospectively recalculated with a model-based approach using Monte Carlo n-Particle Code (MCNP) 6 (Dm,m-MC) and Oncentra Brachy ACE (Dm,m-ACE). Tissue segmentation consisted of assigning uniform compositions and mass densities to predefined Hounsfield Unit (HU) thresholds. Resulting doses were compared according to dose volume histogram parameters typical for clinical routine. These included the percentage liver volume receiving 5 Gy (V5Gy) or 10 Gy (V10Gy), the maximum dose to one cubic centimeter (D1cc) of organs at risk, the clinical target volume (CTV) fractions receiving 150% (V150), 100% (V100), 95% (V95) and 90% (V90) of the prescribed dose and the absolute doses to 95% (D95) and 90% (D90) of the CTV volumes.

RESULTS

Doses from Oncentra Brachy ACE agreed well with MC simulations. Differences were seen far from the source, in low-density regions and bone structures. Median percentage deviations were 1.1% for the liver V5Gy and 0.4% for the liver V10Gy, with deviations of largest magnitude amounting to 2.2% and 1.0%, respectively. Organs at risk had median deviations ranging from 0.3% to 1.5% for D1cc, with outliers ranging up to 4.6%. CTV volume parameter deviations ranged between -1.5% and 0.5%, dose parameter deviations ranged mostly between -2% and 1%, with two outliers at -4.0% and -3.4% for a small CTV.

Down-regulation of lncRNA PCGEM1 inhibits cervical carcinoma by modulating the miR-642a-5p/LGMN axis

LncRNA PCGEM1 (PCGEM1) has been reported to exert essential effects on the development and progress of various tumors, while the detailed effects and possible mechanisms of PCGEM1 in cervical carcinoma remain unknown. In the present study, PCGEM1 was over-expressed in cervical carcinoma cells as evidenced by real-time quantitative polymerase chain reaction (RT-qPCR) assay. Knockdown of PCGEM1 significantly repressed proliferation, migration, and invasion, while induced G1 arrest in cervical carcinoma cells. In addition, PCGEM1 was predicted to target miR-642a-5p by bioinformatics software, which was further confirmed by luciferase reporter assay. Besides, RT-qPCR assay indicated that miR-642a-5p expression was decreased in cervical carcinoma cells and knockdown of PCGEM1 could accelerate miR-642a-5p expression. Moreover, inhibition of miR-642a-5p partly abolished the functions of PCGEM1 knockdown on proliferation, cell cycle, migration and invasion of cervical carcinoma cells. Furthermore, miR-642a-5p could bind to the 3'-UTR of LGMN, which was over-expressed in the cervical carcinoma cells. Suppression of LGMN partly restored the functions of miR-642a-5p inhibitor on proliferation, cell cycle distribution, migration and invasion in the cervical carcinoma cells treated with the PCGEM1 shRNA. Taken together, our data indicated that knockdown of PCGEM1 inhibited proliferation, migration and invasion in cervical carcinoma by modulating the miR-642a-5p/ LGMN axis.

Quantitative analysis of intra-fractional variation in CT-based image guided brachytherapy for cervical cancer patients

We quantified intra-fractional dose variation and organ movement during CT-based 3D-image guided brachytherapy (3D-IGBT) in cervical cancer patients. Fifteen patients who underwent CT-based 3D-IGBT were studied. For all patients, pre-delivery CT for treatment planning after applicator insertion and post-delivery CT after dose delivery without changing the applicator position were acquired. Pre- and post-delivery CT were rigidly fused by matching the inserted applicator and planned dose on pre-delivery CT (pre-delivery dose) was mapped on post-delivery CT (post-delivery dose). D2, D1, and D0.1 cm³ of the rectum and bladder were compared between pre- and post-delivery doses with contours on each CT image. Organ movement and deformation was evaluated using deformation vector fields calculated by deformable image registration between pre- and post-delivery CT. We also evaluated dose variation and DVF between with and without a catheter to control filling. Differences in all DVH parameters were <±3% in physical dose and ± 5% in EQD2. However, a > 15% dose difference was found in 13.8% of the fractions in rectum D2 cm³ and in 11.1% of those in bladder D2 cm³. The mean value of DVF for bladder was larger than that of rectum, especially for the superior-inferior (S-I) direction. Insertion catheters in bladder reduced mean dose and DVF variation compared with that of without catheters. In fraction groups with large dose increasing, DVF in the S-I direction was significantly larger than that of other fraction groups. Our results indicated that preparation is needed to reduce changes in the S-I direction affect dose variation.

The dosimetric impact of replacing the TG-43 algorithm by model based dose calculation for liver brachytherapy

PURPOSE

To compare treatment plans for interstitial high dose rate (HDR) liver brachytherapy with ¹⁹²Ir calculated according to current-standard TG-43U1 protocol with model-based dose calculation following TG-186 protocol.

METHODS

We retrospectively evaluated dose volume histogram (DVH) parameters for liver, organs at risk (OARs) and clinical target volumes (CTVs) of 20 patient cases diagnosed with hepatocellular carcinoma (HCC) or metastatic colorectal cancer (mCRC). Dose calculations on a homogeneous water geometry (TG-43U1 surrogate) and on a computed tomography (CT) based geometry (TG-186) were performed using Monte Carlo (MC) simulations. The CTs were segmented based on a combination of assigning TG-186 recommended tissues to fixed Hounsfield Unit (HU) ranges and using organ contours delineated by physicians. For the liver, V_5Gy and V_10Gy were analysed, and for OARs the dose to 1 cubic centimeter (D_1cc). Target coverage was assessed by calculating V₁₅₀, V₁₀₀, V₉₅ and V₉₀ as well as D₉₅ and D₉₀. For every DVH parameter, median, minimum and maximum values of the deviations of TG-186 from TG-43U1 were analysed.

RESULTS

TG-186-calculated dose was found to be on average lower than dose calculated with TG-43U1. The deviation of highest magnitude for liver parameters was -6.2% of the total liver volume. For OARs, the deviations were all smaller than or equal to -0.5 Gy. Target coverage deviations were as high as -1.5% of the total CTV volume and -3.5% of the prescribed dose.

CONCLUSIONS

In this study we found that TG-43U1 overestimates dose to liver tissue compared to TG-186. This finding may be of clinical importance for cases where dose to the whole liver is the limiting factor.