Applying Self-Supervised Learning to Image Quality Assessment in Chest CT Imaging.

Many new reconstruction techniques have been deployed to allow low-dose CT examinations. Such reconstruction techniques exhibit nonlinear properties, which strengthen the need for a task-based measure of image quality. The Hotelling observer (HO) is the optimal linear observer and provides a lower bound of the Bayesian ideal observer detection performance. However, its computational complexity impedes its widespread practical usage. To address this issue, we proposed a self-supervised learning (SSL)-based model observer to provide accurate estimates of HO performance in very low-dose chest CT images. Our approach involved a two-stage model combining a convolutional denoising auto-encoder (CDAE) for feature extraction and dimensionality reduction and a support vector machine for classification. To evaluate this approach, we conducted signal detection tasks employing chest CT images with different noise structures generated by computer-based simulations. We compared this approach with two supervised learning-based methods: a single-layer neural network (SLNN) and a convolutional neural network (CNN). The results showed that the CDAE-based model was able to achieve similar detection performance to the HO. In addition, it outperformed both SLNN and CNN when a reduced number of training images was considered. The proposed approach holds promise for optimizing low-dose CT protocols across scanner platforms.

Optimized radiological alert thresholds based on device-dosimetric information to predict peak skin dose between 2 and 4 Gy during vascular fluoroscopically guided intervention.

OBJECTIVES: To provide radiologists and physicists with methodological tools to improve patient management after vascular fluoroscopically guided intervention (FGI) by providing optimized thresholds (OT) values that could be used as a surrogate to the thresholds classically proposed by the National Council on Radiation Protection (NCRP) or could be useful to adapt their own substantial radiation dose levels (SRDL) values. METHODS: PSD of 2000-4000 mGy after FGI were calculated for 258 patients with dedicated software. Overall, the kerma and KAP 3D-ROC curves were used to assess the sensitivity (SEN) and specificity (SPE) of NCRP thresholds and OT for each PSD. Kiviat diagram and density curves were plotted for the best SEN/SPE pair of 3D-ROC curves and compared to the NCRP thresholds. RESULTS: OT for both kerma and KAP generating the best SEN/SPE couple for PSD of 2000-4000 mGy were obtained. The SEN/SPE couple of each OT was always better than that obtained using NCRP ones. The best OT among all those calculated providing the highest SEN/SPE values for kerma (3020.5 mGy) and KAP (741.02 Gy.cm2) were obtained when PSD was equal to 3300 mGy. CONCLUSIONS: We have calculated OT in terms of kerma and KAP based on 3D-ROC curves analysis and peak skin dose calculations that can be obtained to better predict high skin dose. The use of OT that predicted PSD greater than 3000 mGy is likely to improve patient follow-up. The methodology developed in this work could be adapted to other institutions in order to better define their own SRDL. KEY POINTS: • Optimized dose thresholds in terms of kerma and KAP based on 3D-ROC curves analysis and peak skin dose calculations between 2000 and 4000 mGy can be obtained to better predict high skin dose. • Patients receiving a peak skin dose between 2000 and 4000 mGy have their follow-up enhanced by using the optimized thresholds instead of the NCRP thresholds. • The best-optimized thresholds, corresponding to 3020.5 mGy and 741.02 Gy.cm2 for kerma and KAP respectively can be used instead of NRCP ones to trigger patient follow-up after fluoroscopically guided vascular interventions.

Comparison of supervised-learning approaches for designing a channelized observer for image quality assessment in CT.

BACKGROUND: The current paradigm for evaluating computed tomography (CT) system performance relies on a task-based approach. As the Hotelling observer (HO) provides an upper bound of observer performances in specific signal detection tasks, the literature advocates HO use for optimization purposes. However, computing the HO requires calculating the inverse of the image covariance matrix, which is often intractable in medical applications. As an alternative, dimensionality reduction has been extensively investigated to extract the task-relevant features from the raw images. This can be achieved by using channels, which yields the channelized-HO (CHO). The channels are only considered efficient when the channelized observer (CO) can approximate its unconstrained counterpart. Previous work has demonstrated that supervised learning-based methods can usually benefit CO design, either for generating efficient channels using partial least squares (PLS) or for replacing the Hotelling detector with machine-learning (ML) methods. PURPOSE: Here we investigated the efficiency of a supervised ML-algorithm used to design a CO for predicting the performance of unconstrained HO. The ML-algorithm was applied either (1) in the estimator for dimensionality reduction, or (2) in the detector function. METHODS: A channelized support vector machine (CSVM) was employed and compared against the CHO in terms of ability to predict HO performances. Both the CSVM and the CHO were estimated with channels derived from the singular value decomposition (SVD) of the system operator, principal component analysis (PCA), and PLS. The huge variety of regularization strategies proposed by CT system vendors for statistical image reconstruction (SIR) make the generalization capability of an observer a key point to consider upfront of implementation in clinical practice. To evaluate the generalization properties of the observers, we adopted a 2-step testing process: (1) achieved with the same regularization strategy (as in the training phase) and (2) performed using different reconstruction properties. We generated simulated- signal-known-exactly/background-known-exactly (SKE/BKE) tasks in which different noise structures were generated using Markov random field (MRF) regularizations using either a Green or a quadratic, function. RESULTS: The CSVM outperformed the CHO for all types of channels and regularization strategies. Furthermore, even though both COs generalized well to images reconstructed with the same regularization strategy as the images considered in the training phase, the CHO failed to generalize to images reconstructed differently whereas the CSVM managed to successfully generalize. Lastly, the proposed CSVM observer used with PCA channels outperformed the CHO with PLS channels while using a smaller training data set. CONCLUSION: These results argue for introducing the supervised-learning paradigm in the detector function rather than in the operator of the channels when designing a CO to provide an accurate estimate of HO performance. The CSVM with PCA channels proposed here could be used as a surrogate for HO in image quality assessment.

Local control and radionecrosis of brain metastases from non-small-cell lung cancer treated by hypofractionated stereotactic radiotherapy: Evaluation of predictive factors.

Background: The objective of our study was to report predictive factors of local control (LC) and radionecrosis (RN) of brain metastases (BM) of non-small cell lung carcinoma (NSCLC) treated by multifractionated stereotactic radiotherapy (MF-SRT) according to French recommendations. Method: From 2012 to 2020, 87 patients with 101 BM were retrospectively included. The median age was 63 years (37-85). GTV was defined using contrast-enhanced T1w MRI and was isotropically extended by 2 mm to form PTV. Mean maximum BM diameter was 24.5 mm (10-46). Patients were treated with dynamic arctherapy from May 2012 to February 2016 and then with VMAT. The total prescribed dose was 23.1 Gy prescribed to the encompassing 70% isodose, in 3 fractions. Results: LC rates at 6 months, 1 year and 2 years was 95.7%, 90.7% and 87.9% respectively. In multivariate analysis, high GTV Dmin (HR = 0.822, p = 0.012) was in favor of better LC whereas a large maximum diameter was predictive of poor LC (HR = 1.124, p = 0.02). GTV Dmin of 27.4 Gy was identified as a discriminant threshold of LC. In case of GTV Dmin ≥ 27.4 Gy, LC at 1 year was 95.3% versus 75.1% with GTV Dmin < 27.4 Gy. Cumulative incidence of RN at 6 months, 1 year and 2 years was 6.3%, 15.4% and 18.1%, respectively. In multivariate analysis, only dyslipidemia was predictive of RN (HR = 2.69, p = 0.03). No dosimetric predictive factor of RN was found in our study. Conclusion: MF-SRT (3x7.7 Gy on 70% isodose line, with PTV = GTV + 2 mm; according to French recommendations) of BM from NSCLC gives high LC rates with acceptable RN rate. A GTV Dmin of at least 27.4 Gy could be proposed to optimize dosimetric objectives. No dosimetric predictive factors of RN were found in this study. However, dyslipidemia was identified as a potential predictive factor of RN.

Impact of iterative reconstruction algorithms on the applicability of Fourier-based detectability index for x-ray CT imaging.

PURPOSE: The increasing application of iterative reconstruction algorithms in clinical computed tomography to improve image quality and reduce radiation dose, elicits strong interest, and needs model observers to optimize CT scanning protocols objectively and efficiently. The current paradigm for evaluating imaging system performance relies on Fourier methods, which presuppose a linear, wide-sense stationary system. Long-range correlations introduced by iterative reconstruction algorithms may narrow the applicability of Fourier techniques. Differences in the implementation of reconstruction algorithms between manufacturers add further complexity. The present work set out to quantify the errors entailed by the use of Fourier methods, which can lead to design decisions that do not correlate with detectability. METHODS: To address this question, we evaluated the noise properties and the detectability index of the ideal linear observer using the spatial approach and the Fourier-based approach. For this purpose, a homogeneous phantom was imaged on two scanners: the Revolution CT (GE Healthcare) and the Somatom Definition AS+ (Siemens Healthcare) at different exposure levels. Images were reconstructed using different strength levels of IR algorithms available on the systems considered: Adaptative Statistical Iterative Reconstruction (ASIR-V) and Sinogram Affirmed Iterative Reconstruction (SAFIRE). RESULTS: Our findings highlight that the spatial domain estimate of the detectability index is higher than the Fourier domain estimate. This trend is found to be dependent on the specific regularization used by IR algorithms as well as the signal to be detected. The eigenanalysis of the noise covariance matrix and of its circulant approximation yields explanation about the evoked trends. In particular, this analysis suggests that the predictive power of the Fourier-based ideal linear observer depends on the ability of each basis analyzed to be relevant to the signal to be detected. CONCLUSION: The applicability of Fourier techniques is dependent on the specific regularization used by IR algorithms. These results argue for verifying the assumptions made when using Fourier methods since Fourier-task-based detectability index does not always correlate with signal detectability.

Relevance of the combination of external beam radiotherapy with the hypoxia-activated prodrug ICF05016 in an experimental model of extraskeletal myxoid chondrosarcoma.

Currently, there is no gold standard treatment for Extraskeletal Myxoid Chondrosarcomas (EMC) making wide margin surgical resection the most effective alternative treatment. Nevertheless, in previous preclinical studies our lab demonstrated the potential of the hypoxia-activated prodrug (HAP) ICF05016 on EMC murine model inoculated with the H-EMC-SS human cell line. The aim of this study was to assess, in vivo, the relevance of the combination of this HAP with External Beam Radiotherapy (EBR). Firstly EMC-bearing mice were treated with 6 Gy or 12 Gy of EBR (single 6 MV photon). Then for combination of HAP and EBR, animals received 6 doses of ICF05016 (46.8 µmol/kg, intravenously) at 4-day intervals, with 6 Gy EBR performed 24 h after the 3rd dose of HAP. Animals were monitored throughout the study for clinical observations (tumour growth, side effects) and survival studies were performed. From tumour samples, PCNA, Ki-67 and p21 expressions were used as markers of proliferation and cell cycle arrest. Statistical significances were determined using Kruskall-Wallis and log rank tests. The radiosensitivity of the EMC model was demonstrated at 12 Gy with significant inhibition of tumour growth. Then, the HAP strategy potentiated EBR efficacy at a lower dose (6 Gy) by improving survival without generating side effects. Thus, results of this study showed the potential interest of ICF05016 for the combination with EBR in the management of EMC.

Significant correlation between gross tumor volume (GTV) D98% and local control in multifraction stereotactic radiotherapy (MF-SRT) for unresected brain metastases.

BACKGROUND: Stereotactic radiotherapy (SRT) should be applied with a biologically effective dose with an α/ß of 12 (BED12) ≥ 40 Gy to reach a 1-year local control (LC) ≥ 70%. The aims of this retrospective study were to report a series of 81 unresected large brain metastases treated with Linac-based multifraction SRT according to the ICRU 91 and to identify predictive factors associated with LC. METHODS: Included in this study were the first 81 brain metastases (BM) consecutively treated with Linac-based volumetric modulated arc therapy (VMAT) multifraction SRT from 2017 to 2019. The prescribed dose was 33 Gy for the GTV and 23.1 Gy (70% isodose line) for the PTV in 3 fractions (3f). Mean BM largest diameter and GTV were 25.1 mm and 7.2 cc respectively. Mean follow-up was 10.2 months. RESULTS: LC was 79.7% and 69.7% at 1 and 2 years respectively. Significant predictive factors of LC were GTV D98% (HR = 0.84, CI 95% = 0.75-0.95, p = 0.004) and adenocarcinoma as the histological type (HR = 0.29, CI 95% = 0.09-0.96, p = 0.042) in univariate and multivariate analysis. A threshold of 29 Gy for GTV D98% was significantly correlated to LC (1-year LC = 91.9% for GTV D98% ≥ 29 Gy vs 69.6% for GTV D98% < 29 Gy (p = 0.030)), corresponding to a BED12 = 52.4 Gy. No tumor progression was observed for a BED12 ≥ 53.4 Gy, corresponding to a GTV D98% ≥ 20 Gy /1f and GTV D98% ≥ 29.4 Gy 3f. Median OS was 15 months. Symptomatic radionecrosis occurred in 4.9% of cases. CONCLUSION: The GTV D98% is a strong reproducible significant predictive factor of LC for brain SRT. Dose prescription should lead to a GTV BED12 98% ≥ 52.4-53.4 Gy to significantly improve LC, corresponding to respectively a GTV D98% ≥ 19.7-20 Gy/1f and 29-29.4 Gy/3f.

Optimized radiological alert thresholds based on device dosimetric information and peak skin dose in vascular fluoroscopically guided intervention.

OBJECTIVES: The National Council on Radiation Protection (NCRP) report no. 168 recommended that during fluoroscopically guided interventions (FGIs), each patient should be monitored when one of the following thresholds is reached: an air kerma > 5 Gy, a kerma area product (KAP) > 500 Gy.cm2, a fluoroscopy time > 60 min, or a peak skin dose (PSD) > 3 Gy. Whereas PSD is the most accurate metric regarding the prevention of radiological risks, it remains the most difficult parameter to assess. We aimed to evaluate the relevance of the other, more accessible metrics and propose new optimized threshold (OT) for improved patient follow-up. METHODS: Overall, 108 patients who underwent FGI in which at least one NCRP threshold was reached and PSD was measured were considered. The correlation between all metrics was assessed using principal component analysis (PCA). ROC curves and the sensitivity/specificity of both NCRP and OT to predict PSD > 3 Gy were evaluated. RESULTS: The PCA shows that FGI can be decomposed with two components based on time and dose variables. Only KAP and kerma were correlated with PSD. The overall sensitivity and specificity of the new OT regarding KAP (67.6/93.0), kerma (97.3/81.7), and time (62.2/62.0) were better compared with NCRP thresholds (97.3/16.9, 40.5/95.4, and 21.6/74.7). CONCLUSIONS: This study shows that fluoroscopy time is not a relevant metric when used to predict PSDs > 3 Gy. By adapting KAP and kerma thresholds to predict PSD over 3 Gy, patient follow-ups following vascular FGI can be improved. KEY POINTS: • In vascular fluoroscopically guided interventions, principal component analysis demonstrates that between fluoroscopy time, KAP, and kerma, only the two last were correlated to the peak skin dose. • Optimized thresholds replacing NRCP ones obtained with ROC curves analysis were 85,451 µGy.cm2, 2938 mGy, and 41 min for KAP, kerma, and fluoroscopy time respectively. • Improvements to trigger patient follow-up after vascular fluoroscopically guided interventions may be obtained by using the optimized thresholds.

Stereotactic Radiosurgery for Vestibular Schwannomas: Reducing Toxicity With 11 Gy as the Marginal Prescribed Dose.

BACKGROUND: Stereotactic radiosurgery (SRS) is a common treatment option for vestibular schwannomas. Historically, a dose de-escalation of the marginal prescribed dose from 16 Gy to 12-13 Gy has been done to limit toxicity without reducing local control (LC). We aimed to retrospectively report outcomes of Linac-based SRS for vestibular schwannomas treated with different doses. METHODS: Included in the study were 97 stage 1 (1%), 2 (56%), 3 (21.5%), and 4 (21.5%) vestibular schwannomas treated with Linac-based (Novalis®) SRS from 1995 to 2019. No margin was added to the GTV to create the PTV. The median marginal prescribed dose was 14 Gy (range: 12-16 Gy) before 2006 and then 11 Gy for all patients (61 pts). Mean tumor volume was 1.96 cm3, i.e., about 1.6 cm in diameter. Mean follow-up was 8.2 years. RESULTS: Following SRS, LC at 3, 5, and 10 years was 100%, 98.4%, and 95.6%, respectively [100% for those with ≤ 13 Gy as the marginal prescribed dose (NS)]. Toxicity to the trigeminal nerve was reported in 7.2% of cases (3.3% and 0% for transient and permanent toxicity for 11 Gy). The marginal prescribed dose was the only significant predictive factor in univariate and multivariate analysis (HR = 1.77, 95% CI = 1.07-3.10, p = 0.028). Toxicity to the facial nerve was reported in 6.2% of cases. The marginal prescribed dose was again the only significant predictive factor in univariate and multivariate analysis (HR = 1.31, 95% CI = 0.77-2.23, p = 0.049). CONCLUSION: Linac-based SRS for stages 1-3 vestibular schwannomas provides excellent outcomes: a 10-year LC rate of over 95%, with a permanent facial or trigeminal toxicity rate of under 5%. A marginal prescribed dose of 11 Gy seems to decrease nerve toxicity and facial toxicity in particular, without reducing LC. Prospective studies with longer follow-up are needed.

Significant Correlation Between Overall Survival and Mean Lung Dose in Lung Stereotactic Body Radiation Therapy (SBRT).

BACKGROUND: After stereotactic body radiation therapy (SBRT) for medically inoperable stage I non-small-cell lung cancer (NSCLC), more patients die of comorbidities, particularly severe pulmonary insufficiency, than of tumor progression. The aim of this study was to evaluate correlation between lung biologically effective dose (BED) with an α/ß ratio of 3 Gy (BED3) and overall survival (OS) for these patients. METHODS: From 2012 to 2017, we have developed a prospectively updated institutional database for all first 100 consecutively treated patients with inoperable Stage 1 (T1T2N0M0) NSCLC. All SBRT were conducted on a Novalis Tx® LINAC with two coplanar dynamic conformal arcs (84%) or with coplanar volumetric modulated arc therapy (VMAT) (16%). Mean GTV and PTV were 8.6 cc and 50.8 cc, respectively. The marginal dose prescribed to the PTV was the 80% isodose line (IDL), i.e., 54 Gy in 3 fractions for 76 patients (BED10 = 126 Gy) and 50 Gy in 5 fractions for 24 patients (BED10 = 83.3 Gy). Pulmonary heterogeneity has been taken into account by using Monte Carlo or AAA algorithms. Median follow-up was 25 months. RESULTS: At 1, 2, 3 and 5 years, local control (LC) was respectively 100, 98.2, 98.2, and 77.7%, and OS was respectively 83, 71.2, 58.1, and 33.2% (median OS was 49 months). Significant OS prognostic factors in univariate and multivariate analysis were mean lung BED3 (HR = 1.14, p = 0.01) and PTV volume (HR = 1.01, p = 0.004). A mean lung BED3 ≤ 5 Gy was significantly associated with a doubling of median OS from 29 months to more than 60 months (not achieved, p = 0.0068). For patients with a forced expiratory volume in 1 second (FEV1) ≤ 40%, a mean lung BED3 ≤ 4 Gy was significantly associated with a doubling of median OS from 23 to 46 months (p = 0.019). CONCLUSION: Mean lung BED3 is strongly and significantly associated with OS in SBRT for inoperable Stage I NSCLC. For all treated patients, a mean lung BED3 ≤ 5 Gy lead to a doubling of median OS. This threshold value should be reduced to 4 Gy for patients with FEV1 ≤ 40%.

Patterns of Failure After Linear Accelerator Radiosurgery for Cerebral Arteriovenous Malformations.

BACKGROUND: Numerous studies have assessed the predictive factors for the arteriovenous malformation (AVM) response to stereotactic radiosurgery (SRS). However, only a few have discussed the causes of failure. The aim of the present study was to evaluate the patterns of failure in patients with AVM who had undergone linear accelerator SRS. METHODS: We performed a retrospective analysis of 288 patients who had undergone linear accelerator SRS in our institution from 1995 to 2011. Failure was defined from the findings of the follow-up angiogram at 5 years, with failure identified in 44 patients. The distribution of causes was estimated using a descriptive analysis of literature-based causes, including a minimal margin dose of <18 Gy, a residual nidus outside the initial targeted volume, previous embolization, recanalization, and the size of the target volume. We also analyzed the associations among the causes. RESULTS: Incomplete nidus identification (41%) and previous embolization (77%) were the most frequently observed conditions in patients with failure. Patients who had undergone previous embolization, for whom the cause of failure had always been identified (P = 0.001), were younger (P = 0.004) and had had a larger nidus volume (P = 0.025). Recanalization was rare (5 of 34 patients) and had occurred exclusively in women (P = 0.048). Larger nidus volumes were less frequent (mean, 2.18 ± 2.2 cm3; range, 0.13-10.8 cm3) and had been observed mainly in women when >2 cm3 (P = 0.012). An insufficient dose was observed in 9 patients and had occurred in the case of a larger volume (P = 0.031), which had resulted in dosimetry constraints in 3 patients and treatment in the vicinity of eloquent zones in 6 patients. No known cause was found in 5 patients, 4 of whom had had a low Spetzler-Martin grade (I and II; P = 0.003), suggestive of radioresistance. CONCLUSION: The results of our detailed analysis have highlighted the distribution of the causes of failure and the potential role of radioresistance in treatment failure.

Second course of stereotactic radiosurgery for locally recurrent brain metastases: Safety and efficacy.

In the present study, we have evaluated the efficacy and toxicity of repeated brain metastases (BM) stereotactic radiosurgery (SRS2) following local failure of a prior radiosurgical procedure (SRS1). Between December 1996 and August 2015, 30 patients with 36 BM underwent SRS2 with a median dose of 18Gy. All BM were located outside critical structures. Following SRS2, local control at 6 months and one year were respectively 82.9% (IC 95%: 67.6-91.9) and 67.8% (IC 95%: 51-81). On multivariate analysis, planning target volume (PTV) < 3cc (HR: 0.19 (0.1-0.52)) and whole brain radiotherapy (WBRT) prior to SRS2 (HR: 0.25 (0.1-0.64)) were significantly associated with a better local control. One- and two-year overall survival rates after SRS2 were respectively 65.5% (IC 95%: 47.3-80%) and 27.6% (IC 95%: 14.7-45.7). Median overall survival following SRS2 was 14.2 months (range 1-106). Nineteen (63%) patients died from progressive systemic disease. Three (10%) patients died from out-field progressive brain disease and 8 (27%) in-field. Concerning toxicities, edema, radionecrosis, and hemorrhages were identified in 5 (12.8%), 4 (10.2%), and 5 (12.8%) patients respectively. No toxicity resulted in a neurological deficit. On univariate analysis, toxicities were significantly associated with PTV > 7cc (p = 0.02) and all patients had a WBRT before SRS2. A second course of SRS for locally recurrent brain metastases showed encouraging rates of local control. This treatment led to acceptable toxicities, especially for brain metastases smaller than 7cc, in our selected cohort of patients with BM located outside critical structures. Further studies are needed.

STAT3 Serine 727 Phosphorylation: A Relevant Target to Radiosensitize Human Glioblastoma.

Radiotherapy is an essential component of glioma standard treatment. Glioblastomas (GBM), however, display an important radioresistance leading to tumor recurrence. To improve patient prognosis, there is a need to radiosensitize GBM cells and to circumvent the mechanisms of resistance caused by interactions between tumor cells and their microenvironment. STAT3 has been identified as a therapeutic target in glioma because of its involvement in mechanisms sustaining tumor escape to both standard treatment and immune control. Here, we studied the role of STAT3 activation on tyrosine 705 (Y705) and serine 727 (S727) in glioma radioresistance. This study explored STAT3 phosphorylation on Y705 (pSTAT3-Y705) and S727 (pSTAT3-S727) in glioma cell lines and in clinical samples. Radiosensitizing effect of STAT3 activation down-modulation by Gö6976 was explored. In a panel of 15 human glioma cell lines, we found that the level of pSTAT3-S727 was correlated to intrinsic radioresistance. Moreover, treating GBM cells with Gö6976 resulted in a highly significant radiosensitization associated to a concomitant pSTAT3-S727 down-modulation only in GBM cell lines that exhibited no or weak pSTAT3-Y705. We report the constitutive activation of STAT3-S727 in all GBM clinical samples. Targeting pSTAT3-S727 mainly in pSTAT3-Y705-negative GBM could be a relevant approach to improve radiation therapy.

Highly efficient radiosensitization of human glioblastoma and lung cancer cells by a G-quadruplex DNA binding compound.

Telomeres are nucleoprotein structures at the end of chromosomes which stabilize and protect them from nucleotidic degradation and end-to-end fusions. The G-rich telomeric single-stranded DNA overhang can adopt a four-stranded G-quadruplex DNA structure (G4). Stabilization of the G4 structure by binding of small molecule ligands enhances radiosensitivity of tumor cells, and this combined treatment represents a novel anticancer approach. We studied the effect of the platinum-derived G4-ligand, Pt-ctpy, in association with radiation on human glioblastoma (SF763 and SF767) and non-small cell lung cancer (A549 and H1299) cells in vitro and in vivo. Treatments with submicromolar concentrations of Pt-ctpy inhibited tumor proliferation in vitro with cell cycle alterations and induction of apoptosis. Non-toxic concentrations of the ligand were then combined with ionizing radiation. Pt-ctpy radiosensitized all cell lines with dose-enhancement factors between 1.32 and 1.77. The combined treatment led to increased DNA breaks. Furthermore, a significant radiosensitizing effect of Pt-ctpy in mice xenografted with glioblastoma SF763 cells was shown by delayed tumor growth and improved survival. Pt-ctpy can act in synergy with radiation for efficient killing of cancer cells at concentrations at which it has no obvious toxicity per se, opening perspectives for future therapeutic applications.

MRI tumor response and clinical outcomes after LINAC radiosurgery on 50 patients with recurrent malignant gliomas.

BACKGROUND: The place of radiosurgery (RS) as an option in the treatment of recurrent malignant glioma is still debated on in the absence of prospective randomized trials. OBJECTIVE: To assess the clinical outcome and MRI response after radiosurgery of recurrent malignant glioma. METHODS: We evaluated 50 consecutive patients treated in a single institution. Between 2001 and 2008, 34 glioblastoma (GBM) and 16 anaplastic oligodendroglioma (AO) patients were treated with linear accelerator (Linac) RS for recurrence. RESULTS: The median marginal dose was 15 Gy and the median gross tumor volume (GTV) was 6.64 ml. No patient had acute (< 3 months) neurological morbidity after RS. Ten patients (20%) were reoperated on after RS, histopathological findings included necrosis alone in 3 cases (6%). The median overall survival was 21.5 months for GBM and 57.9 months for AO. The median survival after RS was 9.5 months for GBM and 32.9 months for AO. The median progression-free survival after RS was 6.7 months for GBM and 18 months for AO. MRI volume modifications after RS showed a transitory reduction or stabilization of disease linked to significantly improved survival in 58.8% of patients with GBM, 81.1% of patients with AO. Pathological subtype (AO versus GBM), MRI response, KPS >70, marginal dose > 13 Gy, largest diameter of GTV < 25 mm and GTV < 7 ml were the main prognostic factors, associated with improved survival or PFS from RS. CONCLUSION: The magnitude of the survival increase compared to historical RPA classes may not be due to selection bias alone. Linac RS in selected patients with recurrent malignant glioma was well tolerated, effective and can be considered as one of several re-treatment options.

Capillary permeability and extracellular volume fraction in uterine cervical cancer as patient outcome predictors: measurements by using dynamic MRI spin-lattice relaxometry.

PURPOSE: To improve the outcome prediction of uterine cervical carcinoma by measuring the vascular permeability (k(ep)) and the extracellular volume fraction (v(e)) of the tumor from Dynamic T(1)- IRM Relaxometry. MATERIALS AND METHODS: Twenty-six patients with proven cervical carcinoma were divided into good outcome and poor outcome groups. Classic tumor prognostic factors, the longest diameter L and the volume V of the tumor, were measured from morphologic MR images. The tumor parameters k(ep) and v(e) were determined from the relaxometry time-curve acquired during the contrast uptake after a bolus intravenous injection of an extracellular contrast agent. RESULTS: All "small" tumors (L<35 mm or V<11 cm(3)) were good outcome with 100% sensitivity but a rather low specificity (36% and 43% for L and V, respectively). With regard to the physiopathological parameter k(ep), "large" tumors (L >or= 35 mm) can also be classified as good outcome on the condition that k(ep) >or= 2.2 min(-1) with 100% sensitivity and 89% specificity. Regarding the extracellular volume fraction (v(e)), no significant difference was observed between the two groups. CONCLUSION: Measurement of the tumor vascular permeability might be useful to predict prognostic, to evaluate the treatment efficacy, and to adapt a proper therapy schedule.

Magnetic resonance imaging measurements of vascular permeability and extracellular volume fraction of breast tumors by dynamic Gd-DTPA-enhanced relaxometry.

Vascular permeability (k(ep), min(-1)) and extracellular volume fraction (v(e)) are tissue parameters of great interest to characterize malignant tumor lesions. Indeed, it is well known that tumors with high blood supply better respond to therapy than poorly vascularized tumors, and tumors with large extracellular volume tend to be more malignant than tumors showing lower extracellular volume. Furthermore, the transport of therapeutic agents depends on both extracellular volume fraction and vessel permeability. Thus, before treatment, these tissue parameters may prove useful to evaluate tumor aggressiveness and to predict responsiveness to therapy and variations during cytotoxic therapies could allow to assess treatment efficacy and early modified therapy schedules in case of poor responsiveness. As a consequence, there is a need to develop methods that could be routinely used to determine these tissue parameters. In this work, blood-tissue permeability and extracellular volume fraction information were derived from magnetic resonance imaging dynamic longitudinal relaxation rate (R(1)) mapping obtained after an intravenous bolus injection of Gd-DTPA in a group of 92 female patients with breast lesions, 68 of these being histologically proven to be with carcinoma. For the sake of comparison, 24 benign lesions were studied. The measurement protocol based on two-dimensional gradient echo sequences and a monoexponential plasma kinetic model was that validated in the occasion of previous animal experiments. As a consequence of neoangiogenesis, results showed a higher permeability in malignant than in benign lesions, whereas the extracellular volume fraction value did not allow any discrimination between benign and malignant lesions. The method, which can be easily implemented whatever the imaging system used, could advantageously be used to quantify lesion parameters (k(ep) and v(e)) in routine clinical imaging. Because of its large reproducibility, the method could be useful for intersite comparisons and follow-up studies.

Magnetic resonance imaging contrast-enhanced relaxometry of breast tumors: an MRI multicenter investigation concerning 100 patients.

Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) using extracellular contrast agents has proved to be useful for the characterization of breast tumors. DCE-MRI has demonstrated a high sensitivity (around 95%) but a rather poor and controversial specificity, varying, according to the different studies, from 45% to 90%. In order to increase (a) the specificity and (b) the robustness of this quantitative approach in multicenter evaluation (five MRI units), a quantitative approach called dynamic relaxometry has been developed. According to the proposed method, the time-dependent longitudinal relaxation rate measured on region of interest of the lesion was calculated during the contrast uptake, after intravenous bolus injection of contrast agent. A specifically developed method was used for fast R(1) measurements. Relaxometry time curves are fitted to the Tofts model allowing the measurement of the parameters describing the enhancement curve (maximum relation rate enhancement, initial, 30-s and 60-s slopes) and the tissue parameters [transfer constant (K(trans) min(-1)) and extracellular extravascular space fraction (v(e))]. Correspondence factorial analysis followed by hierarchical ascendant classification are then performed on the different parameters. Higher K(trans) values were observed in infiltrative ductal carcinomas than in infiltrative lobular carcinomas, in agreement with data published by other groups. Specificity of DCE-MRI has been increased up to 85%, with a sensitivity of 95% with K(trans)/v(e) and enhancement index I (ratio of initial slope by maximum relaxation rate enhancement). A multiparametric data analysis of the calculated parameters opens the way to include quantitative image-based information in new nosologic approaches to breast tumors.