First-in-human study of SBRT and adenosine pathway blockade to potentiate the benefit of immunochemotherapy in early-stage luminal B breast cancer: results of the safety run-in phase of the Neo-CheckRay trial.

BACKGROUND: Luminal B breast cancer (BC) presents a worse prognosis when compared with luminal A BC and exhibits a lower sensitivity to chemotherapy and a lower immunogenicity in contrast to non-luminal BC subtypes. The Neo-CheckRay clinical trial investigates the use of stereotactic body radiation therapy (SBRT) directed to the primary tumor in combination with the adenosine pathway inhibitor oleclumab to improve the response to neo-adjuvant immuno-chemotherapy in luminal B BC. The trial consists of a safety run-in followed by a randomized phase II trial. Here, we present the results of the first-in-human safety run-in. METHODS: The safety run-in was an open-label, single-arm trial in which six patients with early-stage luminal B BC received the following neo-adjuvant regimen: paclitaxel q1w×12 → doxorubicin/cyclophosphamide q2w×4; durvalumab (anti-programmed cell death receptor ligand 1 (PD-L1)) q4w×5; oleclumab (anti-CD73) q2w×4 → q4w×3 and 3×8 Gy SBRT to the primary tumor at week 5. Surgery must be performed 2-6 weeks after primary systemic treatment and adjuvant therapy was given per local guidelines, RT boost to the tumor bed was not allowed. Key inclusion criteria were: luminal BC, Ki67≥15% or histological grade 3, MammaPrint high risk, tumor size≥1.5 cm. Primary tumor tissue samples were collected at three timepoints: baseline, 1 week after SBRT and at surgery. Tumor-infiltrating lymphocytes, PD-L1 and CD73 were evaluated at each timepoint, and residual cancer burden (RCB) was calculated at surgery. RESULTS: Six patients were included between November 2019 and March 2020. Median age was 53 years, range 37-69. All patients received SBRT and underwent surgery 2-4 weeks after the last treatment. After a median follow-up time of 2 years after surgery, one grade 3 adverse event (AE) was reported: pericarditis with rapid resolution under corticosteroids. No grade 4-5 AE were documented. Overall cosmetical breast evaluation after surgery was 'excellent' in four patients and 'good' in two patients. RCB results were 2/6 RCB 0; 2/6 RCB 1; 1/6 RCB 2 and 1/6 RCB 3. CONCLUSIONS: This novel treatment combination was considered safe and is worth further investigation in a randomized phase II trial. TRIAL REGISTRATION NUMBER: NCT03875573.

New Explainable Deep CNN Design for Classifying Breast Tumor Response Over Neoadjuvant Chemotherapy.

PURPOSE: To reduce breast tumor size before surgery, Neoadjuvant Chemotherapy (NAC) is applied systematically to patients with local breast cancer. However, with the existing clinical protocols, it is not yet possible to have an early prediction of the effect of chemotherapy on a breast tumor. Predicting the response to chemotherapy could reduce toxicity and delay effective treatment. Computational analysis of Dynamic Contrast-Enhanced Magnetic Resonance Images (DCE-MRI) through Deep Convolution Neural Network (CNN) has proved a significant performance in classifying responders and no responder's patients. This study intends to present a new explainable Deep Learning (DL) model predicting the breast cancer response to chemotherapy based on multiple MRI inputs. MATERIAL AND METHODS: In this study, a cohort of 42 breast cancer patients who underwent chemotherapy was used to train and validate the proposed DL model. This dataset was provided by the Jules Bordet institute of radiology in Brussels, Belgium. 14 external subjects were used to validate the DL model to classify responding or non-responding patients on temporal DCE-MRI volumes. The model performance was assessed by the Area Under the receiver operating characteristic Curve (AUC), accuracy, and features map visualization according to pathological complete response (Ground truth). RESULTS: The developed deep learning architecture was able to predict the responding breast tumors to chemotherapy treatment in the external validation dataset with an AUC of 0.93 using parallel learning MRI images acquired at different moments. The visual results showed that the most important extracted features from non-responding tumors are in the peripheral and external tumor regions. The model proposed in this study is more efficient compared to those proposed in the literature. CONCLUSION: Even with a limited training dataset size, the developed multi-input CNN model using DCE-MR images acquired before and following the first chemotherapy was able to predict responding and non-responding tumors with higher accuracy. Thanks to the visualization of the extracted characteristics by the DL model on the responding and non-responding tumors, the latter could be used henceforth in clinical analysis after its evaluation based on more extra data.

Multi-input deep learning architecture for predicting breast tumor response to chemotherapy using quantitative MR images.

PURPOSE: Neoadjuvant chemotherapy (NAC) aims to minimize the tumor size before surgery. Predicting response to NAC could reduce toxicity and delays to effective intervention. Computational analysis of dynamic contrast-enhanced magnetic resonance images (DCE-MRI) through deep convolution neural network (CNN) has shown a significant performance to distinguish responders and no responder's patients. This study intends to present a new deep learning (DL) model predicting the breast cancer response to NAC based on multiple MRI inputs. METHODS: A cohort of 723 axial slices extracted from 42 breast cancer patients who underwent NAC therapy was used to train and validate the developed DL model. This dataset was provided by our collaborator institute of radiology in Brussels. Fourteen external cases were used to validate the best obtained model to predict pCR based on pre- and post-chemotherapy DCE-MRI. The model performance was assessed by area under the receiver operating characteristic curve (AUC), accuracy, sensitivity, specificity, and feature map visualization. RESULTS: The developed multi-inputs deep learning architecture was able to predict the pCR to NAC treatment in the validation dataset with an AUC of 0.91 using combined pre- and post-NAC images. The visual results showed that the most important extracted features from non-pCR tumors are in the peripheral region. The proposed method was more productive than the previous ones. CONCLUSION: Even with a limited training dataset size, the proposed and developed CNN model using DCE-MR images acquired before and after the first chemotherapy was able to classify pCR and non-pCR patients with substantial accuracy. This model could be used hereafter in clinical analysis after its evaluation based on more extra data.

Prognostic value of a three-scale grading system based on combining molecular imaging with 68Ga-DOTATATE and 18F-FDG PET/CT in patients with metastatic gastroenteropancreatic neuroendocrine neoplasias.

We investigated on the added prognostic value of a three-scale combined molecular imaging with 68Ga-DOTATATE and 18F-FDG PET/CT, (compared to Ki-67 based histological grading), in gastroenteropancreatic neuroendocrine neoplasia patients. 85 patients with histologically proven metastatic gastroenteropancreatic neuroendocrine neoplasias, who underwent combined PET/CT imaging were retrospectively evaluated. Highest Ki-67 value available at time of 18F-FDG PET/CT was recorded. Patients were classified according to World Health Organization/European Neuroendocrine Tumor Society histological grades (G1, G2, G3) and into three distinct imaging categories (C1: all lesions are 18F-FDG negative/68Ga-DOTATATE positive, C2: patients with one or more 18F-FDG positive lesions, all of them 68Ga-DOTATATE positive, C3: patients with one or more 18F-FDG positive lesions, at least one of them 68Ga-DOTATATE negative). The primary endpoint of the study was Progression-Free Survival, assessed from the date of 18F-FDG PET/CT to the date of radiological progression according to Response Evaluation Criteria In Solid Tumors version 1.1. Classification according to histological grade did not show significant statistical difference in median Progression-Free Survival between G1 and G2 but was significant between G2 and G3 patients. In contrast, median Progression-Free Survival was significantly higher in C1 compared to C2 and in C2 compared to C3 patients, revealing three distinctive imaging categories, each with highly distinctive prognosis. Our three-scale combined 68Ga-DOTATATE/18F-FDG PET imaging classification holds high prognostic value in patients with metastatic gastroenteropancreatic neuroendocrine neoplasias.

Early prediction of neoadjuvant treatment outcome in locally advanced breast cancer using parametric response mapping and radial heterogeneity from breast MRI.

BACKGROUND: Early prediction of nonresponse is essential in order to avoid inefficient treatments. PURPOSE: To evaluate if parametrical response mapping (PRM)-derived biomarkers could predict early morphological response (EMR) and pathological complete response (pCR) 24-72 hours after initiation of chemotherapy treatment and whether concentric analysis of nonresponding PRM regions could better predict response. STUDY TYPE: This was a retrospective analysis of prospectively acquired cohort, nonrandomized, monocentric, diagnostic study. POPULATION: Sixty patients were initially recruited, with 39 women participating in the final cohort. FIELD STRENGTH/SEQUENCE: A 1.5T scanner was used for MRI examinations. ASSESSMENT: Dynamic contrast-enhanced (DCE)-MR images were acquired at baseline (timepoint 1, TP1), 24-72 hours after the first chemotherapy (TP2), and after the end of anthracycline treatment (TP3). PRM was performed after fusion of T1 subtraction images from TP1 and TP2 using an affine registration algorithm. Pixels with an increase of more than 10% of their value (PRMdce+) were corresponding nonresponding regions of the tumor. Patients with a decrease of maximum diameter (%dDmax) between TP1 and TP3 of more than 30% were defined as EMR responders. pCR patients achieved a residual cancer burden score of 0. STATISTICAL TESTS: T-test, receiver operating characteristic (ROC) curves, and logistic regression were used for the analysis. RESULTS: PRM showed a statistical difference between pCR response groups (P < 0.01) and AUC of 0.88 for the prediction of non-pCR. Logistic regression analysis demonstrated that PRMdce+ and Grade II were significant (P < 0.01) for non-pCR prediction (AUC = 0.94). Peripheral tumor region demonstrated higher performance for the prediction of non-pCR (AUC = 0.85) than intermediate and central zones; however, statistical comparison showed no significant difference. DATA CONCLUSION: PRM could be predictive of non-pCR 24-72 hours after initiation of chemotherapy treatment. Moreover, the peripheral region showed increased AUC for non-pCR prediction and increased signal intensity during treatment for non-pCR tumors, information that could be used for optimal tissue sampling. LEVEL OF EVIDENCE: 1 Technical Efficacy Stage: 4 J. Magn. Reson. Imaging 2020;51:1403-1411.

The impact of breast MRI workup on tumor size assessment and surgical planning in patients with early breast cancer.

BACKGROUND: The size and focality of the primary tumor in breast cancer (BC) influence therapeutic decision making. The purpose of this study was to evaluate whether preoperative breast magnetic resonance imaging (MRI) is helpful for the assessment of tumor size and surgical planning in early BC. METHODS: We performed a retrospective review of a prospectively collected database of 174 patients treated at a single institution for invasive BC who had complete documentation of the tumor size from mammography (MMG), ultrasonography (US), and MRI. RESULTS: A total of 186 breast tumors were analyzed. Mean tumor size varied by imaging method: 14.7 mm by MMG, 13.8 mm by US, and 17.9 mm by MRI. The concordance between breast imaging techniques (BIT) and final pathology with a cutoff ≤ 2 mm was 34.8% for MRI, 32.1% for US, and 27.2% for MMG. US and MMG underestimated while MRI and MMG overestimated the real tumor size. Concordance was the same in premenopausal women for MRI and US at 35%, while concordance was higher in postmenopausal women for MRI. Correlations between size determined by BIT and histopathological size were best with MRI (0.59), compared to US (0.56) or MMG (0.42). Intrinsic subtypes of BC had different concordances according to imaging method, but no significant associations were found. MRI examination revealed additional lesions in 13.8% of patients, 69% of these lesions were malignant. MRI changed the surgical plan in 15 patients (8.6%), and the rate of mastectomy increased by 6.9%. CONCLUSIONS: MRI estimates BC tumor size more accurately than US or MMG, but a significant overestimation exists. Complementary MRI examination improved the concordance for tumor size between BIT and final pathology in 16.7%. MRI did not alter surgical planning for most patients and allowed more appropriate treatment for 8% of them.

A PRM approach for early prediction of breast cancer response to chemotherapy based on registered MR images.

PURPOSE: This study aims to provide and optimize a performing algorithm for predicting the breast cancer response rate to the first round of chemotherapy using Magnetic Resonance Imaging (MRI). This provides an early recognition of breast tumor reaction to chemotherapy by using the Parametric Response Map (PRM) method. METHODS: PRM may predict the breast cancer response to chemotherapy by analyzing voxel-by-voxel temporal intra-tumor changes during one round of chemotherapy. Indeed, the tumor recognizes intra-tumor changes concerning its vascularity, which is an important criterion in the present study. This method is mainly based on spatial image affine registration between the breast tumor MRI volumes, acquired before and after the first cycle of chemotherapy, and region growing segmentation of the tumor volume. To evaluate our method, we used a retrospective study of 40 patients provided by a collaborating institute. RESULTS: PRM allows a color map to be created with the percentages of positive, negative and stable breast tumor response during the first round of chemotherapy, identifying each region with its response rate. We assessed the accuracy of the proposed method using technical and medical validation methods. The technical validation was based on landmarks-based registration and fully manual segmentation. The medical evaluation was based on the accuracy calculation of the standard reference of anatomic pathology. The p-values and the Area Under the Curve (AUC) of the Receiver Operating Characteristics were calculated to evaluate the proposed PRM method. CONCLUSION: We performed and evaluated the proposed PRM method to study and analyze the behavior of a tumor during the first round of chemotherapy, based on the intra-tumor changes of MR breast tumor images. The AUC obtained for the PRM method is considered as relevant in the early prediction of breast tumor response.

Total choline quantification measured by 1H MR spectroscopy as early predictor of response after neoadjuvant treatment for locally advanced breast cancer: The impact of immunohistochemical status.

BACKGROUND: Validation of new biomarkers is essential for the early evaluation of neoadjuvant treatments. PURPOSE: To determine whether measurements of total choline (tCho) by 1H spectroscopy could predict morphological or pathological complete response (pCR) of neoadjuvant treatment and whether breast cancer subgroups are related to prediction accuracy. STUDY TYPE: Prospective, nonrandomized, monocentric, diagnostic study. POPULATION: Sixty patients were initially included with 39 women participating in the final cohort. FIELD STRENGTH/SEQUENCE: A 1.5T scanner was used for acquisition and MRS was performed using the syngo GRACE sequence. ASSESSMENT: MRS and MRI examinations were performed at baseline (TP1), 24-72 hours after first chemotherapy (TP2), after the end of anthracycline treatment (TP3), and MRI only after the end of taxane treatment (TP4). Early (EMR) and late (LMR) morphological response were defined as %ΔDmax13 or %ΔDmax14, respectively. Responders were patients with %ΔDmax >30. Pathological complete response (pCR) patients achieved a residual cancer burden score of 0. STATISTICAL TESTS: T-test, receiver operating characteristic (ROC) curves, multiple regression, logistic regression, one-way analysis of variance (ANOVA) analysis were used for the analysis. RESULTS: At TP1 there was a significant difference between response groups for tCho1 concerning EMR prediction (P = 0.05) and pCR (P < 0.05) and for Kep 1 (P = 0.03) concerning LMR prediction. At TP2, no modification of tCho and other parameters could predict response. At TP3, ΔtCho, ΔDmax, and ΔVol could predict LMR (P < 0.05 for all parameters), pCR (P < 0.05 for all parameters), and ΔKtrans could predict only pCR (P = 0.04). Logistic regression at baseline showed the highest area under the curve (AUC) of 0.9 for prediction of pCR. The triple negative (TN) subgroup showed significantly higher tCho at baseline (P = 0.02) and higher ΔtCho levels at TP3 (P < 0.05). DATA CONCLUSION: Baseline measurements of tCho in combination with clinicopathological criteria could predict non-pCR with a high AUC. Furthermore, tCho quantification for prediction of pCR was more sensitive for TN tumors. LEVEL OF EVIDENCE: 1 Technical Efficacy: Stage 4 J. Magn. Reson. Imaging 2018;48:982-993.

Quantitative DCE-MRI for prediction of pathological complete response following neoadjuvant treatment for locally advanced breast cancer: the impact of breast cancer subtypes on the diagnostic accuracy.

OBJECTIVES: To assess whether DCE-MRI pharmacokinetic (PK) parameters obtained before and during chemotherapy can predict pathological complete response (pCR) differently for different breast cancer groups. METHODS: Eighty-four patients who received neoadjuvant chemotherapy for locally advanced breast cancer were retrospectively included. All patients underwent two DCE-MRI examinations, one before (EX1) and one during treatment (EX2). Tumours were classified into different breast cancer groups, namely triple negative (TNBC), HER2+ and ER+/HER2-, and compared with the whole population (WP). PK parameters Ktrans and Ve were extracted using a two-compartment Tofts model. RESULTS: At EX1, Ktrans predicted pCR for WP and TNBC. At EX2, maximum diameter (Dmax) predicted pCR for WP and ER+/HER2-. Both PK parameters predicted pCR in WP and TNBC and only Ktrans for the HER2+. pCR was predicted from relative difference (EX1 - EX2)/EX1 of Dmax and both PK parameters in the WP group and only for Ve in the TNBC group. No PK parameter could predict response for ER+/HER-. ROC comparison between WP and breast cancer groups showed higher but not statistically significant values for TNBC for the prediction of pCR CONCLUSIONS: Quantitative DCE-MRI can better predict pCR after neoadjuvant treatment for TNBC but not for the ER+/HER2- group. KEY POINTS: • DCE-MRI-derived pharmacokinetic parameters can predict response status of neoadjuvant chemotherapy treatment. • Ktrans can better predict pCR for the triple negative group. • No pharmacokinetic parameter could predict response for the ER+/HER2- group.

Computer-aided detection of colonic polyps using low-dose CT acquisitions.

RATIONALE AND OBJECTIVES: This report proposes an alternative method for the automatic detection of colonic polyps that is robust enough to be directly applicable on low-dose computed tomographic data. MATERIALS AND METHODS: The polyp modeling process takes into account both the gray-level appearance of polyps (intensity profiles) and their geometry (extended Gaussian images). Spherical harmonic decompositions are used for comparison purposes, allowing fast estimation of the similarity between a candidate and a set of previously computed models. Starting from the original raw data (acquired at 55 mA), five patient data sets (prone and supine scans) are reconstructed at different dose levels (to 5 mA) by using different kernel filters, slice overlaps, and increments. Additionally, the efficacy of applying an edge-preserving smoothing filter before detection is assessed. RESULTS: Although image quality decreases when decreasing acquisition milliamperes, all polyps greater than 6 mm are detected successfully, even at 15 mA. Although not important at high doses, smoothing improves detection results for ultra-low-dose (tube current<15 mA) data. CONCLUSION: The advantage of low-dose scans is a significant decrease in effective dose from 4.93 to 1.61 mSv while retaining high detection values, particularly important when thinking of population screening.

Removal of Plaque and Stent Artifacts in Subtraction CT Angiography Using Nonrigid Registration and a Volume Penalty.

Computed tomography angiography (CTA) is an established tool for vascular imaging. However, high-intense structures in the contrast image can seriously hamper luminal visualisation. This can be solved by subtraction CTA, where a native image is subtracted from the contrast image. However, patient and organ motion limit the application of this technique. Within this paper, a fully automated intensity-based nonrigid 3D registration algorithm for subtraction CT angiography is presented, using a penalty term to avoid volume change during registration. Visual and automated validation on four clinical datasets clearly show that the algorithm strongly reduces motion artifacts in subtraction CTA. Most artifacts disappear, also artifacts caused by minimal displacement of stents or calcified plaques, allowing a 2D and 3D artifact-free visualisation of the vessel lumen. This enables a quick overview of the whole vascular structure and opens the possibility to the visualisation of smaller vessels.

Computer aided detection for low-dose CT colonography.

The paper describes a method for automatic detection of colonic polyps, robust enough to be directly applied to low-dose CT colonographic datasets. Polyps are modeled using gray level intensity profiles and extended Gaussian images. Spherical harmonic decompositions ensure an easy comparison between a polyp candidate and a set of polypoid models, found in a previously built database. The detection sensitivity and specificity values are evaluated at different dose levels. Starting from the original raw-data (acquired at 55mAs), 5 patient datasets (prone and supine scans) are reconstructed at different dose levels (down to 5mAs), using different kernel filters and slice increments. Although the image quality decreases when lowering the acquisition mAs, all polyps above 6mm are successfully detected even at 15 mAs. Accordingly the effective dose can be reduced from 4.93mSv to 1.61 mSv, without affecting detection capabilities, particularly important when thinking of population screening.

Plaque and stent artifact reduction in subtraction CT angiography using nonrigid registration and a volume penalty.

Computed tomography angiography (CTA) is an established tool for vessel imaging. Yet, high-intense structures in the contrast image can seriously hamper luminal visualisation. This can be solved by subtraction CTA, where a native image is subtracted from the contrast image. However, patient and organ motion limit the application of this technique. Within this paper, a fully automated intensity-based nonrigid 3D registration algorithm for subtraction CT angiography is presented, using a penalty term to avoid volume change during registration. Visual and automated validation on four clinical datasets clearly show that the algorithm strongly reduces motion artifacts in subtraction CTA. With our method, 39% to 99% of the artifacts disappear, also those caused by minimal displacement of stents or calcified plaques. This results in a better visualisation of the vessel lumen, also of the smaller vessels, allowing a faster and more accurate inspection of the whole vascular structure, especially in case of stenosis.