MIDOS: a novel stochastic model towards a treatment planning system for microsphere dosimetry in liver tumors.

PURPOSE: Transarterial radioembolization (TARE) procedures treat liver tumors by injecting radioactive microspheres into the hepatic artery. Currently, there is a critical need to optimize TARE towards a personalized dosimetry approach. To this aim, we present a novel microsphere dosimetry (MIDOS) stochastic model to estimate the activity delivered to the tumor(s), normal liver, and lung. METHODS: MIDOS incorporates adult male/female liver computational phantoms with the hepatic arterial, hepatic portal venous, and hepatic venous vascular trees. Tumors can be placed in both models at user discretion. The perfusion of microspheres follows cluster patterns, and a Markov chain approach was applied to microsphere navigation, with the terminal location of microspheres determined to be in either normal hepatic parenchyma, hepatic tumor, or lung. A tumor uptake model was implemented to determine if microspheres get lodged in the tumor, and a probability was included in determining the shunt of microspheres to the lung. A sensitivity analysis of the model parameters was performed, and radiation segmentectomy/lobectomy procedures were simulated over a wide range of activity perfused. Then, the impact of using different microspheres, i.e., SIR-Sphere®, TheraSphere®, and QuiremSphere®, on the tumor-to-normal ratio (TNR), lung shunt fraction (LSF), and mean absorbed dose was analyzed. RESULTS: Highly vascularized tumors translated into increased TNR. Treatment results (TNR and LSF) were significantly more variable for microspheres with high particle load. In our scenarios with 1.5 GBq perfusion, TNR was maximum for TheraSphere® at calibration time in segmentectomy/lobar technique, for SIR-Sphere® at 1-3 days post-calibration, and regarding QuiremSphere® at 3 days post-calibration. CONCLUSION: This novel approach is a decisive step towards developing a personalized dosimetry framework for TARE. MIDOS assists in making clinical decisions in TARE treatment planning by assessing various delivery parameters and simulating different tumor uptakes. MIDOS offers evaluation of treatment outcomes, such as TNR and LSF, and quantitative scenario-specific decisions.

Gadolinium-Based Nanoparticles Sensitize Ovarian Peritoneal Carcinomatosis to Targeted Radionuclide Therapy.

Ovarian cancer (OC) is the most lethal gynecologic malignancy (5-y overall survival rate, 46%). OC is generally detected when it has already spread to the peritoneal cavity (peritoneal carcinomatosis). This study investigated whether gadolinium-based nanoparticles (Gd-NPs) increase the efficacy of targeted radionuclide therapy using [177Lu]Lu-DOTA-trastuzumab (an antibody against human epidermal growth factor receptor 2). Gd-NPs have radiosensitizing effects in conventional external-beam radiotherapy and have been tested in clinical phase II trials. Methods: First, the optimal activity of [177Lu]Lu-DOTA-trastuzumab (10, 5, or 2.5 MBq) combined or not with 10 mg of Gd-NPs (single injection) was investigated in athymic mice bearing intraperitoneal OC cell (human epidermal growth factor receptor 2-positive) tumor xenografts. Next, the therapeutic efficacy and toxicity of 5 MBq of [177Lu]Lu-DOTA-trastuzumab with Gd-NPs (3 administration regimens) were evaluated. NaCl, trastuzumab plus Gd-NPs, and [177Lu]Lu-DOTA-trastuzumab alone were used as controls. Biodistribution and dosimetry were determined, and Monte Carlo simulation of energy deposits was performed. Lastly, Gd-NPs' subcellular localization and uptake, and the cytotoxic effects of the combination, were investigated in 3 cancer cell lines to obtain insights into the involved mechanisms. Results: The optimal [177Lu]Lu-DOTA-trastuzumab activity when combined with Gd-NPs was 5 MBq. Moreover, compared with [177Lu]Lu-DOTA-trastuzumab alone, the strongest therapeutic efficacy (tumor mass reduction) was obtained with 2 injections of 5 mg of Gd-NPs/d (separated by 6 h) at 24 and 72 h after injection of 5 MBq of [177Lu]Lu-DOTA-trastuzumab. In vitro experiments showed that Gd-NPs colocalized with lysosomes and that their radiosensitizing effect was mediated by oxidative stress and inhibited by deferiprone, an iron chelator. Exposure of Gd-NPs to 177Lu increased the Auger electron yield but not the absorbed dose. Conclusion: Targeted radionuclide therapy can be combined with Gd-NPs to increase the therapeutic effect and reduce the injected activities. As Gd-NPs are already used in the clinic, this combination could be a new therapeutic approach for patients with ovarian peritoneal carcinomatosis.

The complexity of DNA damage by radiation follows a Gamma distribution: insights from the Microdosimetric Gamma Model.

Introduction: DNA damage is the main predictor of response to radiation therapy for cancer. Its Q8 quantification and characterization are paramount for treatment optimization, particularly in advanced modalities such as proton and alpha-targeted therapy. Methods: We present a novel approach called the Microdosimetric Gamma Model (MGM) to address this important issue. The MGM uses the theory of microdosimetry, specifically the mean energy imparted to small sites, as a predictor of DNA damage properties. MGM provides the number of DNA damage sites and their complexity, which were determined using Monte Carlo simulations with the TOPAS-nBio toolkit for monoenergetic protons and alpha particles. Complexity was used together with a illustrative and simplistic repair model to depict the differences between high and low LET radiations. Results: DNA damage complexity distributions were were found to follow a Gamma distribution for all monoenergetic particles studied. The MGM functions allowed to predict number of DNA damage sites and their complexity for particles not simulated with microdosimetric measurements (yF) in the range of those studied. Discussion: Compared to current methods, MGM allows for the characterization of DNA damage induced by beams composed of multi-energy components distributed over any time configuration and spatial distribution. The output can be plugged into ad hoc repair models that can predict cell killing, protein recruitment at repair sites, chromosome aberrations, and other biological effects, as opposed to current models solely focusing on cell survival. These features are particularly important in targeted alpha-therapy, for which biological effects remain largely uncertain. The MGM provides a flexible framework to study the energy, time, and spatial aspects of ionizing radiation and offers an excellent tool for studying and optimizing the biological effects of these radiotherapy modalities.

A systematic analysis of the particle irradiation data ensemble in the key of the microdosimetric kinetic model: Should clonogenic data be used for clinical relative biological effectiveness?

PURPOSE: To perform a systematic analysis of the Particle Irradiation Data Ensemble (PIDE) database for clonogenic survival assays in the context of the Microdosimetric Kinetic Model (MKM). METHODS AND MATERIAL: Our study used data from the PIDE database containing data on various cell lines and radiation types. Two main parameters of the MKM were determined experiment-wise: the domain radius, which accounts for the increase of the linear parameter as a function of LET or lineal energy, and the nucleus radius, which accounts for the overkilling effect at LET high enough. We used experiments with LET less and more than 75 keV/µm to determine domain and nucleus radius, respectively. Experiments with cells in asynchronous phase of the cell cycle and monoenergetic beams were considered, and data from 294 out of 461 available experiments with protons, alpha, and carbon beams were used. RESULTS: Domain and nucleus radii were determined for 32 cell lines as the median among cell-specific experiments after filtering experiments using protons, α-particles, and carbon ions, including 28 human cells and 12 rodent cells. The median values found for domain radii were 380 nm for normal human cells, 390 nm for tumor human cells, 295 nm for normal rodent cells, and 525 nm for tumor rodent cells (only one experiment with rodent tumor cells) with large variability across cell lines and across experiments on each cell line. CONCLUSIONS: Large inter-experiment variabilities were found for the same cell lines, based on enormous experimental uncertainties and different experimental conditions. Our analysis raises questions about how convenient is to use clonogenic data to feed RBE models to be utilized in the clinical practice in particle therapy.

Absence of Tissue-Sparing Effects in Partial Proton FLASH Irradiation in Murine Intestine.

Ultra-high dose rate irradiation has been reported to protect normal tissues more than conventional dose rate irradiation. This tissue sparing has been termed the FLASH effect. We investigated the FLASH effect of proton irradiation on the intestine as well as the hypothesis that lymphocyte depletion is a cause of the FLASH effect. A 16 × 12 mm2 elliptical field with a dose rate of ~120 Gy/s was provided by a 228 MeV proton pencil beam. Partial abdominal irradiation was delivered to C57BL/6j and immunodeficient Rag1-/-/C57 mice. Proliferating crypt cells were counted at 2 days post exposure, and the thickness of the muscularis externa was measured at 280 days following irradiation. FLASH irradiation did not reduce the morbidity or mortality of conventional irradiation in either strain of mice; in fact, a tendency for worse survival in FLASH-irradiated mice was observed. There were no significant differences in lymphocyte numbers between FLASH and conventional-dose-rate mice. A similar number of proliferating crypt cells and a similar thickness of the muscularis externa following FLASH and conventional dose rate irradiation were observed. Partial abdominal FLASH proton irradiation at 120 Gy/s did not spare normal intestinal tissue, and no difference in lymphocyte depletion was observed. This study suggests that the effect of FLASH irradiation may depend on multiple factors, and in some cases dose rates of over 100 Gy/s do not induce a FLASH effect and can even result in worse outcomes.

Large anatomical changes in head-and-neck cancers - A dosimetric comparison of online and offline adaptive proton therapy.

Purpose: This work evaluates an online adaptive (OA) workflow for head-and-neck (H&N) intensity-modulated proton therapy (IMPT) and compares it with full offline replanning (FOR) in patients with large anatomical changes. Methods: IMPT treatment plans are created retrospectively for a cohort of eight H&N cancer patients that previously required replanning during the course of treatment due to large anatomical changes. Daily cone-beam CTs (CBCT) are acquired and corrected for scatter, resulting in 253 analyzed fractions. To simulate the FOR workflow, nominal plans are created on the planning-CT and delivered until a repeated-CT is acquired; at this point, a new plan is created on the repeated-CT. To simulate the OA workflow, nominal plans are created on the planning-CT and adapted at each fraction using a simple beamlet weight-tuning technique. Dose distributions are calculated on the CBCTs with Monte Carlo for both delivery methods. The total treatment dose is accumulated on the planning-CT. Results: Daily OA improved target coverage compared to FOR despite using smaller target margins. In the high-risk CTV, the median D98 degradation was 1.1 % and 2.1 % for OA and FOR, respectively. In the low-risk CTV, the same metrics yield 1.3 % and 5.2 % for OA and FOR, respectively. Smaller setup margins of OA reduced the dose to all OARs, which was most relevant for the parotid glands. Conclusion: Daily OA can maintain prescription doses and constraints over the course of fractionated treatment, even in cases of large anatomical changes, reducing the necessity for manual replanning in H&N IMPT.

Proton FLASH effects on mouse skin at different oxygen tensions.

Objective. Irradiation at FLASH dose rates (>40 Gy s-1) has received great attention due to its reported normal tissue sparing effect. The FLASH effect was originally observed in electron irradiations but has since been shown to also occur with both photon and proton beams. Several mechanisms have been proposed to explain the tissue sparing at high dose rates, including effects involving oxygen, such as depletion of oxygen within the irradiated cells. In this study, we investigated the protective role of FLASH proton irradiation on the skin when varying the oxygen concentration.Approach. Our double scattering proton system provided a 1.2 × 1.6 cm2elliptical field at a dose rate of ∼130 Gy s-1. The conventional dose rate was ∼0.4 Gy s-1. The legs of the FVB/N mice were marked with two tattooed dots and fixed in a holder for exposure. To alter the skin oxygen concentration, the mice were breathing pure oxygen or had their legs tied to restrict blood flow. The distance between the two dots was measured to analyze skin contraction over time.Main results. FLASH irradiation mitigated skin contraction by 15% compared to conventional dose rate irradiation. The epidermis thickness and collagen deposition at 75 d following 25 to 30 Gy exposure suggested a long-term protective function in the skin from FLASH irradiation. Providing the mice with oxygen or reducing the skin oxygen concentration removed the dose-rate-dependent difference in response.Significance. FLASH proton irradiation decreased skin contraction, epidermis thickness and collagen deposition compared to standard dose rate irradiations. The observed oxygen-dependence of the FLASH effect is consistent with, but not conclusive of, fast oxygen depletion during the exposure.

Impact of DNA Geometry and Scoring on Monte Carlo Track-Structure Simulations of Initial Radiation-Induced Damage.

Track structure Monte Carlo simulations are a useful tool to investigate the damage induced to DNA by ionizing radiation. These simulations usually rely on simplified geometrical representations of the DNA subcomponents. DNA damage is determined by the physical and physicochemical processes occurring within these volumes. In particular, damage to the DNA backbone is generally assumed to result in strand breaks. DNA damage can be categorized as direct (ionization of an atom part of the DNA molecule) or indirect (damage from reactive chemical species following water radiolysis). We also consider quasi-direct effects, i.e., damage originated by charge transfers after ionization of the hydration shell surrounding the DNA. DNA geometries are needed to account for the damage induced by ionizing radiation, and different geometry models can be used for speed or accuracy reasons. In this work, we use the Monte Carlo track structure tool TOPAS-nBio, built on top of Geant4-DNA, for simulation at the nanometer scale to evaluate differences among three DNA geometrical models in an entire cell nucleus, including a sphere/spheroid model specifically designed for this work. In addition to strand breaks, we explicitly consider the direct, quasi-direct, and indirect damage induced to DNA base moieties. We use results from the literature to determine the best values for the relevant parameters. For example, the proportion of hydroxyl radical reactions between base moieties was 80%, and between backbone, moieties was 20%, the proportion of radical attacks leading to a strand break was 11%, and the expected ratio of base damages and strand breaks was 2.5-3. Our results show that failure to update parameters for new geometric models can lead to significant differences in predicted damage yields.

Modified Geometry of 106Ru Asymmetric Eye Plaques to Improve Dosimetric Calculations in Ophthalmic Brachytherapy.

Ru/Rh asymmetric plaques for ophthalmic brachytherapy have special geometric designs with a cutout intended to prevent irradiation of critical ocular structures proximal to the tumor. In this work, we present new geometric models for PENELOPE+PenEasy Monte Carlo simulations of these applicators, differing from the vendor-reported geometry, that better match their real geometry to assess their dosimetric impact. Simulation results were benchmarked to experimental dosimetric data from radiochromic film measurements, data provided by the manufacturer in the calibration certificates, and other experimental results published in the literature, obtaining, in all cases, better agreement with the modified geometries. The clinical impact of the new geometric models was evaluated by simulating real clinical cases using patient-specific eye models. The cases calculated using the modified geometries presented higher doses to the critical structures proximal to the cutout region. The modified geometric models presented in this work provide a more accurate representation of the asymmetric plaques, greatly improving the agreement between Monte Carlo calculations and experimental measurements. Lack of consideration of accurate geometric models has been shown to be translated into notable increases in dose to organs at risk in clinical cases.

Correlation of LET With MRI Changes in Brain and Potential Implications for Normal Tissue Complication Probability for Patients With Meningioma Treated With Pencil Beam Scanning Proton Therapy.

PURPOSE: This study aimed to investigate the correlation between imaging changes in brain normal tissue and the spatial distribution of linear energy transfer (LET) for a cohort of patients with meningioma treated with scanned proton beams. Then, assuming imaging changes are induced by cell lethality, we studied the correlation between normal tissue complication probability and LET. METHODS AND MATERIALS: Magnetic resonance imaging T2/fluid attenuated inversion recovery acquired at different intervals after proton radiation were coregistered with the planning computed tomography (CT) images from 26 patients with meningioma with abnormalities after proton radiation therapy. For this purpose, the T2/fluid attenuated inversion recovery areas not on the original magnetic resonance images were contoured, and the LET values for each voxel in the patient geometry were calculated to investigate the correlation between the position of imaging changes and the LET at those positions. To separate the effect of the dose as the inductor of these changes, we compared the LET in these areas with a sample of voxels matching the dose distributions across the image change areas. Patients with a higher LET in image change areas were grouped to verify whether they shared common characteristics. RESULTS: Eleven of the patients showed higher dose-averaged LET (LETd) in imaging change regions than in the group of voxels with the same dose. This group of patients had significantly shallower targets for their treatment than the other 15 and used fewer beams and angles. CONCLUSIONS: This study points toward the possibility that areas with imaging change are more likely to occur in regions with high dose or in areas with lower dose but increased LETd. The effect of LETd on imaging changes seems to be more relevant when treating superficial lesions with few nonopposed beams. However, most patients did not show a spatial correlation between their image changes and the LETd values, limiting the cases for the possible role of high LET as a toxicity inductor.

Pre- and post-treatment image-based dosimetry in90Y-microsphere radioembolization using the TOPAS Monte Carlo toolkit.

Objective. To evaluate the pre-treatment and post-treatment imaging-based dosimetry of patients treated with 90Y-microspheres, including accurate estimations of dose to tumor, healthy liver and lung. To do so, the Monte Carlo (MC) TOPAS platform is in this work extended towards its utilization in radionuclide therapy.Approach. Five patients treated at the Massachusetts General Hospital were selected for this study. All patients had data for both pre-treatment SPECT-CT imaging using 99mTc-MAA as a surrogate of the 90Y-microspheres treatment and SPECT-CT imaging immediately after the 90Y activity administration. Pre- and post-treatment doses were computed with TOPAS using the SPECT images to localize the source positions and the CT images to account for tissue inhomoegeneities. We compared our results with analytical calculations following the voxel-based MIRD scheme.Main results. TOPAS results largely agreed with the MIRD-based calculations in soft tissue regions: the average difference in mean dose to the liver was 0.14 Gy GBq-1(2.6%). However, dose distributions in the lung differed considerably: absolute differences in mean doses to the lung ranged from 1.2 to 6.3 Gy GBq-1and relative differences from 153% to 231%. We also found large differences in the intra-hepatic dose distributions between pre- and post-treatment imaging, but only limited differences in the pulmonary dose.Significance. Doses to lung were found to be higher using TOPAS with respect to analytical calculations which may significantly underestimate dose to the lung, suggesting the use of MC methods for 90Y dosimetry. According to our results, pre-treatment imaging may still be representative of dose to lung in these treatments.

The relation between microdosimetry and induction of direct damage to DNA by alpha particles.

In radiopharmaceutical treatmentsα-particles are employed to treat tumor cells. However, the mechanism that drives the biological effect induced is not well known. Being ionizing radiation,α-particles can affect biological organisms by producing damage to the DNA, either directly or indirectly. Following the principle that microdosimetry theory accounts for the stochastic way in which radiation deposits energy in sub-cellular sized volumes via physical collisions, we postulate that microdosimetry represents a reasonable framework to characterize the statistical nature of direct damage induction byα-particles to DNA. We used the TOPAS-nBio Monte Carlo package to simulate direct damage produced by monoenergetic alpha particles to different DNA structures. In separate simulations, we obtained the frequency-mean lineal energy (yF) and dose-mean lineal energy (yD) of microdosimetric distributions sampled with spherical sites of different sizes. The total number of DNA strand breaks, double strand breaks (DSBs) and complex strand breaks per track were quantified and presented as a function of eitheryForyD.The probability of interaction between a track and the DNA depends on how the base pairs are compacted. To characterize this variability on compactness, spherical sites of different size were used to match these probabilities of interaction, correlating the size-dependent specific energy (z) with the damage induced. The total number of DNA strand breaks per track was found to linearly correlate withyFandzFwhen using what we defined an effective volume as microdosimetric site, while the yield of DSB per unit dose linearly correlated withyDorzD,being larger for compacted than for unfolded DNA structures. The yield of complex breaks per unit dose exhibited a quadratic behavior with respect toyDand a greater difference among DNA compactness levels. Microdosimetric quantities correlate with the direct damage imparted on DNA.

Modelling Dose Effects from Space Irradiations: Combination of High-LET and Low-LET Radiations with a Modified Microdosimetric Kinetic Model.

The Microdosimetric Kinetic Model (MKM) to predict the effects of ionizing radiation on cell colonies is studied and reformulated for the case of high-linear energy transfer (LET) radiations with a low dose. When the number of radiation events happening in a subnuclear domain follows a Poisson distribution, the MKM predicts a linear-quadratic (LQ) survival curve. We show that when few events occur, as for high-LET radiations at doses lower than the mean specific energy imparted to the nucleus, zF,n, a Poisson distribution can no longer be assumed and an initial pure linear relationship between dose and survival fraction should be observed. Predictions of survival curves for combinations of high-LET and low-LET radiations are produced under two assumptions for their comparison: independent and combined action. Survival curves from previously published articles of V79 cell colonies exposed to X-rays, α particles, Ar-ions, Fe-ions, Ne-ions and mixtures of X-rays and each one of the ions are predicted according to the modified MKM. We conclude that mixtures of high-LET and low-LET radiations may enhance the effect of individual actions due to the increase of events in domains provided by the low-LET radiation. This hypothesis is only partially validated by the analyzed experiments.

Radiobiological effectiveness difference of proton arc beams versus conventional proton and photon beams.

This paper aims to demonstrate the difference in biological effectiveness of proton monoenergetic arc therapy (PMAT) compared to intensity modulated proton therapy (IMPT) and conventional 6 MV photon therapy, and to quantify this difference when exposing cells of different radiosensitivity to the same experimental conditions for each modality. V79, H1299 and H460 cells were cultured in petri dishes placed in the central axis of a cylindrical and homogeneous solid water phantom of 20 cm in diameter. For the PMAT plan, cells were exposed to 13 mono-energetic proton beams separated every 15° over a 180° arc, designed to deliver a uniform dose of higher LET to the petri dishes. For the IMPT plans, 3 fields were used, where each field was modulated to cover the full target. Cells were also exposed to 6 MV photon beams in petri dishes to characterize their radiosensitivity. The relative biological effectiveness of the PMAT plans compared with those of IMPT was measured using clonogenic assays. Similarly, in order to study the quantity and quality of the DNA damage induced by the PMAT plans compared to that of IMPT and photons, γ-H2AX assays were conducted to study the relative amount of DNA damage induced by each modality, and their repair rate over time. The clonogenic assay revealed similar survival levels to the same dose delivered with IMPT or x-rays. However, a systematic average of up to a 43% increase in effectiveness in PMAT plans was observed when compared with IMPT. In addition, the repair kinetic assays proved that PMAT induces larger and more complex DNA damage (evidenced by a slower repair rate and a larger proportion of unrepaired DNA damage) than IMPT. The repair kinetics of IMPT and 6 MV photon therapy were similar. Mono-energetic arc beams offer the possibility of taking advantage of the enhanced LET of proton beams to increase TCP. This study presents initial results based on exposing cells with different radiosensitivity to other modalities under the same experimental conditions, but more extensive clonogenic and in-vivo studies will be required to confirm the validity of these results.

Calculation of clinical dose distributions in proton therapy from microdosimetry.

PURPOSE: To introduce a new algorithm-MicroCalc-for dose calculation by modeling microdosimetric energy depositions and the spectral fluence at each point of a particle beam. Proton beams are considered as a particular case of the general methodology. By comparing the results obtained against Monte Carlo computations, we aim to validate the microdosimetric formalism presented here and in previous works. MATERIALS AND METHODS: In previous works, we developed a function on the energy for the average energy imparted to a microdosimetric site per event and a model to compute the energetic spectrum at each point of the patient image. The number of events in a voxel is estimated assuming a model in which the voxel is completely filled by microdosimetric sites. Then, dose at every voxel is computed by integrating the average energy imparted per event multiplied by the number of events per energy beam of the spectral distribution within the voxel. Our method is compared with the proton convolution superposition (PCS) algorithm implemented in Eclipse™ and the fast Monte Carlo code MCsquare, which is here considered the benchmark, for in-water calculations, using in both cases clinically validated beam data. Two clinical cases are considered: a brain and a prostate case. RESULTS: For a SOBP beam in water, the mean difference at the central axis found for MicroCalc is of 0.86% against 1.03% for PCS. Three-dimensional gamma analyses in the PTVs compared with MCsquare for criterion (3%, 3 mm) provide gamma index of 95.07% with MicroCalc vs 94.50% with PCS for the brain case and 99.90% vs 100.00%, respectively, for the prostate case. For selected organs at risk in each case (brainstem and rectum), mean and maximum difference with respect to MCsquare dose are analyzed. In the brainstem, mean differences are 0.25 Gy (MicroCalc) vs 0.56 Gy (PCS), whereas for the rectum, these values are 0.05 Gy (MicroCalc) vs 0.07 Gy (PCS). CONCLUSIONS: The accuracy of MicroCalc seems to be, at least, not inferior to PCS, showing similar or better agreement with MCsquare in the considered cases. Additionally, the algorithm enables simultaneous computation of other quantities of interest. These results seem to validate the microdosimetric methodology in which the algorithm is based on.

Monte Carlo verification of radiotherapy treatments with CloudMC.

BACKGROUND: A new implementation has been made on CloudMC, a cloud-based platform presented in a previous work, in order to provide services for radiotherapy treatment verification by means of Monte Carlo in a fast, easy and economical way. A description of the architecture of the application and the new developments implemented is presented together with the results of the tests carried out to validate its performance. METHODS: CloudMC has been developed over Microsoft Azure cloud. It is based on a map/reduce implementation for Monte Carlo calculations distribution over a dynamic cluster of virtual machines in order to reduce calculation time. CloudMC has been updated with new methods to read and process the information related to radiotherapy treatment verification: CT image set, treatment plan, structures and dose distribution files in DICOM format. Some tests have been designed in order to determine, for the different tasks, the most suitable type of virtual machines from those available in Azure. Finally, the performance of Monte Carlo verification in CloudMC is studied through three real cases that involve different treatment techniques, linac models and Monte Carlo codes. RESULTS: Considering computational and economic factors, D1_v2 and G1 virtual machines were selected as the default type for the Worker Roles and the Reducer Role respectively. Calculation times up to 33 min and costs of 16 € were achieved for the verification cases presented when a statistical uncertainty below 2% (2σ) was required. The costs were reduced to 3-6 € when uncertainty requirements are relaxed to 4%. CONCLUSIONS: Advantages like high computational power, scalability, easy access and pay-per-usage model, make Monte Carlo cloud-based solutions, like the one presented in this work, an important step forward to solve the long-lived problem of truly introducing the Monte Carlo algorithms in the daily routine of the radiotherapy planning process.