Prediction Of Pathologic Response Research Articles

Prediction of pathological response grading for esophageal squamous carcinoma after neoadjuvant chemoradiotherapy based on MRI imaging using PDX.

To confirm the efficacy of magnetic resonance-diffusion weighted imaging (MR-DWI) in esophageal squamous cell carcinoma (ESCC) early pathological response prediction and assessment to neoadjuvant chemoradiotherapy (nCRT) using patient-derived xenografts (PDXs). PDX-bearing mice were randomly divided into two groups: the experimental group receiving cisplatin combined with radiotherapy, whereas the control group receiving normal saline. MRI scans were performed in treatment groups in the before, middle, and end of treatment. The correlations between tumor volumes, ADC values and tumor pathological response at different time nodes were explored. Then, expression of proliferation marker and apoptotic marker were detected using immunohistochemistry, and apoptosis rate was detected by TUNEL assay to further verify the results observed in the PDX models. The ADC values of the experimental group were significantly higher than the control group in the both middle and end stage of treatment (all P< 0.001), however, significant difference was only observed in tumor volume at the end stage of treatment (P< 0.001). Furthermore, the △ADCmid-pre in our study may able to identify tumors with or without pCR to nCRT at an early stage, due to these changes were prior to the changes of tumor volume after treatment. Finally, TUNEL results also showed that the apoptosis rate of the experiment groups increased the most in the middle stage of treatment, especially the groups with pCR, but the highest apoptosis rate occurred in the end of the treatment. Further, the two PDX models with pCR exhibited the highest levels of apoptotic marker (Bax), and lowest levels of proliferation marker (PCNA and Ki-67) in the both middle and end stage of the treatment. ADC values could be used to determine the tumor's response to nCRT, especially in the middle stages of treatment and before the tumor tissue morphology changes, and further, the ADC values were consistent with the potential biomarkers reflecting histopathological changes. Therefore, we suggest that radiation oncologists could refer to the ADC values in the middle stages of treatment when predicting the tumor histopathological response to n CRT in patients with ESCC.

The Role of Multimodal Imaging in Pathological Response Prediction of Locally Advanced Cervical Cancer Patients Treated by Chemoradiation Therapy Followed by Radical Surgery.

This study aimed to develop predictive models for pathological residual disease after neoadjuvant chemoradiation (CRT) in locally advanced cervical cancer (LACC) by integrating parameters derived from transvaginal ultrasound, MRI and PET/CT imaging at different time points and time intervals. Patients with histologically proven LACC, stage IB2-IVA, were prospectively enrolled. For each patient, the three examinations were performed before, 2 and 5 weeks after treatment ("baseline", "early" and "final", respectively). Multivariable logistic regression models to predict complete vs. partial pathological response (pR) were developed and a cost analysis was performed. Between October 2010 and June 2014, 88 patients were included. Complete or partial pR was found in 45.5% and 54.5% of patients, respectively. The two most clinically useful models in pR prediction were (1) using percentage variation of SUVmax retrieved at PET/CT "baseline" and "final" examination, and (2) including high DWI signal intensity (SI) plus, ADC, and SUVmax collected at "final" evaluation (area under the curve (95% Confidence Interval): 0.80 (0.71-0.90) and 0.81 (0.72-0.90), respectively). The percentage variation in SUVmax in the time interval before and after completing neoadjuvant CRT, as well as DWI SI plus ADC and SUVmax obtained after completing neoadjuvant CRT, could be used to predict residual cervical cancer in LACC patients. From a cost point of view, the use of MRI and PET/CT is preferable.

FDG-PET/CT response and prediction of pathological response to neoadjuvant nivolumab and ipilimumab for clinical stage III melanoma.

9585 Background: Neoadjuvant immunotherapy with nivolumab 3mg/kg and ipilimumab 1 mg/kg (N3+I1) for two cycles for clinical stage III melanoma have shown rates of major pathological response (MPR) of approximately 60%.The prognosis of this group seems to be excellent so far. On the other hand, patients classified as pathological non-responders have a worse outcome and early identification of this group may allow us to tailor the treatment before surgery. Methods: We conducted a multicenter retrospective analysis of patients with clinical stage III melanoma treated with neoadjuvant N3+I1 for two cycles who did baseline and pre-operative 18F-FDG-PET/CT. The total number of FGD avid lesions and the percentual difference between the maximum SUV per lesion was calculated. The pathological results were correlated to FGD-PET/CT findings. Results: Between January 2019 and January 2023, 28 patients with clinical stage III melanoma treated with N3+I1 who had baseline and preoperative 18F-FDG PET were identified. Gender: 20 (71%) males, median age (range):55 (34-78), BRAFV600E/K mut: 15(53%), positive nodes on baseline PET/CT:1 = 26 (92%), 2 = 1(4%), and 3 = 1(4%). All known lesions identified by CT scan were also captured by FGD-PET/CT. Site of node(s): axilla = 12 (43%), cervical = 10 (36%), and inguinal = 6(21%). All but one patient received 2 cycles of N3+IP1(1 patient had grade 3 toxicity and received only one cycle). Pathological response: MPR = 19(68%), non-MPR: 8 (28%), and 1(4%) did not undergo surgery due to widespread progression. An increase in maximum SUV and/or appearance of new lesion(s), n = 8(28%), was correlated to non-MPR or metastatic disease in all cases, including a patient who developed sarcoidosis-like reaction with increase of SUV in the index lesion (+68%) and appearance of inflammatory mediastinal lymph-nodes. The median increase was (range): +73% (+11% to 483%). Reduction or stable maximum SUV with no appearance of new lesion, n = 20 (72%), was associated with MPR in 19 patients (95%). One patient,who had G3 colitis (with the need of infliximab), had 32% of reduction in maximum SUV but 100% of viable tumor cells on pathological report. The median decrease (range) was -76% (-1.2% to -100%). Conclusions: FGD-PET/CT may help to predict pathological response in patients with clinical stage III melanoma who undergo to neoadjuvant nivolumab and ipilimumab.

MO-0059 Prediction of pathological response to chemo-radiotherapy in rectal cancer using federated learning

MO-0059 Prediction of pathological response to chemo-radiotherapy in rectal cancer using federated learning

Role of (F-18) Fluorodeoxyglucose Positron Emission Tomography/Computed Tomography in the Prediction of Response to Neoadjuvant Therapy in Esophageal Cancer: Correlation with Pathological Response and Survival.

The purpose of this study is to assess the correlation between metabolic response with fluorodeoxyglucose positron emission tomography/computed tomography (FDG-PET/CT) and pathological response in patients with locally advanced esophageal cancer treated with neoadjuvant chemoradiotherapy and to study FDG-PET parameters for the prediction of pathological response and outcome. Twenty-five patients with locally advanced esophageal cancer underwent two FDG-PET/CT scans for initial staging and after neoadjuvant chemoradiotherapy. FDG uptake in the primary tumor was calculated in both scans (SUVmax, SULpeak, and TLG). Metabolic response was assessed according to the reduction of PET parameters: complete response (mCR = 100%), partial response (mPR ≥50%), and no response (mNR ≤50%). Pathological response was also classified as complete (pCR), partial (pPR), or no response (pNR). Patients were followed up (range, 8-99 months) determining free-disease interval (FDI) and overall survival (OS). Two patients were excluded due to exitus for nonesophageal-related causes. The metabolic response was observed in 18/23 remaining patients (3mCR, 15 mPR), of which 12/18 patients showed a pathological response (3 pCR, 9 pPR). A major discrepancy was observed in 2 mNR patients who achieved pPR. FDI and OS were longer in patients with metabolic response than nonresponders, but no statistical difference was found. No significant correlation was found between PET parameters and pathological response, FDI, and OS. FDG-PET/CT is a useful technique to assess response to neoadjuvant chemoradiotherapy in esophageal cancer. Although in this preliminary study, no correlation between metabolic and pathologic response was found and no statistical differences between responders and nonresponders were observed, a tendency of longer FDI and OS was apparently found in responders patients.

Machine Learning-Based Radiomic Features on Pre-Ablation MRI as Predictors of Pathologic Response in Patients with Hepatocellular Carcinoma Who Underwent Hepatic Transplant

The aim was to investigate the role of pre-ablation tumor radiomics in predicting pathologic treatment response in patients with early-stage hepatocellular carcinoma (HCC) who underwent liver transplant. Using data collected from 2005-2015, we included adult patients who (1) had a contrast-enhanced MRI within 3 months prior to ablation therapy and (2) underwent liver transplantation. Demographics were obtained for each patient. The treated hepatic tumor volume was manually segmented on the arterial phase T1 MRI images. A vector with 112 radiomic features (shape, first-order, and texture) was extracted from each tumor. Feature selection was employed through minimum redundancy and maximum relevance using a training set. A random forest model was developed based on top radiomic and demographic features. Model performance was evaluated by ROC analysis. SHAP plots were constructed in order to visualize feature importance in model predictions. Ninety-seven patients (117 tumors, 31 (32%) microwave ablation, 66 (68%) radiofrequency ablation) were included. The mean model for end-stage liver disease (MELD) score was 10.5 ± 3. The mean follow-up time was 336.2 ± 179 days. Complete response on pathology review was achieved in 62% of patients at the time of transplant. Incomplete pathologic response was associated with four features: two first-order and two GLRM features using univariate logistic regression analysis (p < 0.05). The random forest model included two radiomic features (diagnostics maximum and first-order maximum) and four clinical features (pre-procedure creatinine, pre-procedure albumin, age, and gender) achieving an AUC of 0.83, a sensitivity of 82%, a specificity of 67%, a PPV of 69%, and an NPV of 80%. Pre-ablation MRI radiomics could act as a valuable imaging biomarker for the prediction of tumor pathologic response in patients with HCC.

Pretreatment ultrasound-based deep learning radiomics model for the early prediction of pathologic response to neoadjuvant chemotherapy in breast cancer.

To investigate the predictive performance of the deep learning radiomics (DLR) model integrating pretreatment ultrasound imaging features and clinical characteristics for evaluating therapeutic response after neoadjuvant chemotherapy (NAC) in patients with breast cancer. A total of 603 patients who underwent NAC were retrospectively included between January 2018 and June 2021 from three different institutions. Four different deep convolutional neural networks (DCNNs) were trained by pretreatment ultrasound images using annotated training dataset (n = 420) and validated in a testing cohort (n = 183). Comparing the predictive performance of these models, the best one was selected for image-only model structure. Furthermore, the integrated DLR model was constructed based on the image-only model combined with independent clinical-pathologic variables. Areas under the curve (AUCs) of these models and two radiologists were compared by using the DeLong method. As the optimal basic model, Resnet50 achieved an AUC and accuracy of 0.879 and 82.5% in the validation set. The integrated DLR model, yielding the highest classification performance in predicting response to NAC (AUC 0.962 and 0.939 in the training and validation cohort), outperformed the image-only model and the clinical model and also performed better than two radiologists' prediction (all p < 0.05). In addition, predictive efficacy of the radiologists was improved under the assistance of the DLR model significantly. The pretreatment US-based DLR model could hold promise as a clinical guidance for predicting NAC response of patients with breast cancer, thereby providing benefit of timely treatment strategy adjustment to potential poor NAC responders. • Multicenter retrospective study showed that deep learning radiomics (DLR) model based on pretreatment ultrasound image and clinical parameter achieved satisfactory prediction of tumor response to neoadjuvant chemotherapy (NAC) in breast cancer. • The integrated DLR model could become an effective tool to guide clinicians in identifying potential poor pathological responders before chemotherapy. • The predictive efficacy of the radiologists was improved under the assistance of the DLR model.

Metabolic tumour and nodal response to neoadjuvant chemotherapy on FDG PET-CT as a predictor of pathological response and survival in patients with oesophageal adenocarcinoma

Objectives2-deoxy-2[18F]Fluoro-d-glucose (FDG) PET-CT has an emerging role in assessing response to neoadjuvant therapy in oesophageal cancer. This study evaluated FDG PET-CT in predicting pathological tumour response (pTR), pathological nodal response (pNR) and survival.MethodsCohort study of 75 patients with oesophageal or oesophago-gastric junction (GOJ) adenocarcinoma treated with neoadjuvant chemotherapy then surgery at Guy’s and St Thomas’ NHS Foundation Trust, London (2017–2020). Standardised uptake value (SUV) metrics on pre- and post-treatment FDG PET-CT in the primary tumour (mTR) and loco-regional lymph nodes (mNR) were derived. Optimum SUVmax thresholds for predicting pathological response were identified using receiver operating characteristic analysis. Predictive accuracy was compared to PERCIST (30% SUVmax reduction) and MUNICON (35%) criteria. Survival was assessed using Cox regression.ResultsOptimum tumour SUVmax decrease for predicting pTR was 51.2%. A 50% cut-off predicted pTR with 73.5% sensitivity, 69.2% specificity and greater accuracy than PERCIST or MUNICON (area under the curve [AUC] 0.714, PERCIST 0.631, MUNICON 0.659). Using a 30% SUVmax threshold, mNR predicted pNR with high sensitivity but low specificity (AUC 0.749, sensitivity 92.6%, specificity 57.1%, p = 0.010). pTR, mTR, pNR and mNR were independent predictive factors for survival (pTR hazard ratio [HR] 0.10 95% confidence interval [CI] 0.03–0.34; mTR HR 0.17 95% CI 0.06–0.48; pNR HR 0.17 95% CI 0.06–0.54; mNR HR 0.13 95% CI 0.02–0.66).ConclusionsMetabolic tumour and nodal response predicted pTR and pNR, respectively, in patients with oesophageal or GOJ adenocarcinoma. However, currently utilised response criteria may not be optimal. pTR, mTR, pNR and mNR were independent predictors of survival.Key Points• FDG PET-CT has an emerging role in evaluating response to neoadjuvant therapy in patients with oesophageal cancer.• Prospective cohort study demonstrated that metabolic response in the primary tumour and lymph nodes was predictive of pathological response in a cohort of patients with adenocarcinoma of the oesophagus or oesophago-gastric junction treated with neoadjuvant chemotherapy followed by surgical resection.• Patients who demonstrated a response to neoadjuvant chemotherapy in the primary tumour or lymph nodes on FDG PET-CT demonstrated better survival and reduced rates of tumour recurrence.

Prediction of pathological response after neoadjuvant chemotherapy using baseline FDG PET heterogeneity features in breast cancer

Complete pathological response to neoadjuvant systemic treatment (NAST) in some subtypes of breast cancer (BC) has been used as a surrogate of long-term outcome. The possibility of predicting BC pathological response to NAST based on the baseline 18F-Fluorodeoxyglucose positron emission tomography (FDG PET), without the need of an interim study, is a focus of recent discussion. This review summarises the characteristics and results of the available studies regarding the potential impact of heterogeneity features of the primary tumour burden on baseline FDG PET in predicting pathological response to NAST in BC patients. Literature search was conducted on PubMed database and relevant data from each selected study were collected. A total of 13 studies were eligible for inclusion, all of them published over the last 5 years. Eight out of 13 analysed studies indicated an association between FDG PET-based tumour uptake heterogeneity features and prediction of response to NAST. When features associated with predicting response to NAST were derived, these varied between studies. Therefore, definitive reproducible findings across series were difficult to establish. This lack of consensus may reflect the heterogeneity and low number of included series. The clinical relevance of this topic justifies further investigation about the predictive role of baseline FDG PET.

Multimodality Imaging Review of HER2-positive Breast Cancer and Response to Neoadjuvant Chemotherapy

Human epidermal growth factor receptor 2 (HER2/neu or ErbB2)-positive breast cancers comprise 15%-20% of all breast cancers. The most common manifestation of HER2-positive breast cancer at mammography or US is an irregular mass with spiculated margins that often contains calcifications; at MRI, HER2-positive breast cancer may appear as a mass or as nonmass enhancement. HER2-positive breast cancers are often of intermediate to high nuclear grade at histopathologic analysis, with increased risk of local recurrence and metastases and poorer overall prognosis. However, treatment with targeted monoclonal antibody therapies such as trastuzumab and pertuzumab provides better local-regional control and leads to improved survival outcome. With neoadjuvant treatments, including monoclonal antibodies, taxanes, and anthracyclines, women are now potentially able to undergo breast conservation therapy and sentinel lymph node biopsy versus mastectomy and axillary lymph node dissection. Thus, the radiologist's role in assessing the extent of local-regional disease and response to neoadjuvant treatment at imaging is important to inform surgical planning and adjuvant treatment. However, assessment of treatment response remains difficult, with the potential for different imaging modalities to result in underestimation or overestimation of disease to varying degrees when compared with surgical pathologic analysis. In particular, the presence of calcifications at mammography is especially difficult to correlate with the results of pathologic analysis after chemotherapy. Breast MRI findings remain the best predictor of pathologic response. The authors review the initial manifestations of HER2-positive tumors, the varied responses to neoadjuvant chemotherapy, and the challenges in assessing residual cancer burden through a multimodality imaging review with pathologic correlation. © RSNA, 2023 Quiz questions for this article are available through the Online Learning Center.

Multiparametric MRI-based radiomics nomogram for early prediction of pathological response to neoadjuvant chemotherapy in locally advanced gastric cancer.

To build and validate a multi-parametric MRI (mpMRI)-based radiomics nomogram for early prediction of treatment response to neoadjuvant chemotherapy (NAC) in locally advanced gastric cancer. Baseline MRI were retrospectively enrolled from 141 patients with gastric adenocarcinoma who received NAC followed by radical gastrectomy. According to pathologic response of tumor regression grading (TRG), patients were labeled as responders (TRG = 0 + 1) and non-responders (TRG = 2 + 3) and further divided into a training (n = 85) and validation dataset (n = 56). Radiomics score (Radscore) were built from T2WI, ADC, and venous phase of dynamic-contrasted-enhanced MR imaging. Clinical information, laboratory indicators, MRI parameters, and follow-up data were also recorded. According to multivariable regression analysis, an mpMRI radiomics nomogram was built and its predictive ability was evaluated by ROC analysis. Decision curve analysis was applied to evaluate the clinical usefulness. Kaplan-Meier survival curves based on the nomogram were used to estimate the progression-free survival (PFS) and overall survival (OS) in the validation dataset. Both single sequence-based Radscores and mpMRI radiomics nomogram were associated with pathologic response (p < 0.001). The nomogram achieved the highest diagnostic ability with area under ROC curve of 0.844 (95% CI, 0.749-0.914) and 0.820 (95% CI, 0.695-0.910) in the training and validation datasets. The hazard ratio of the nomogram for PFS and OS prediction was 2.597 (95% CI: 1.046-6.451, log-rank p = 0.023) and 2.570 (95% CI: 1.166-5.666, log-rank p = 0.011). The mpMRI-based radiomics nomogram showed preferable performance in predicting pathologic response to NAC in LAGC. • This study investigated the value of multi-parametric MRI-based radiomics in predicting pathologic response to neoadjuvant chemotherapy in locally advanced gastric cancer. • The nomogram incorporating T2WI Radscore, ADC Radscore, and DCE Radscore as well as Borrmann classification outperformed the single sequence-based Radscore. • The nomogram also exhibited a promising prognostic ability for patient survival and enriched radiomics studies in gastric cancer.

OGC P20 Metabolic tumour and nodal response to neoadjuvant chemotherapy on FDG PET-CT as a predictor of pathological response and survival in patients undergoing surgical resection for locally advanced oesophageal adenocarcinoma

Abstract Background 18F-fluorodeoxyglucose positron emission tomography-computed tomography (FDG PET-CT) is routinely used for staging of oesophageal cancer and has an emerging role in assessing response to neoadjuvant therapy. Several studies have demonstrated that a reduction in FDG avidity of the primary tumour and loco-regional lymph nodes following neoadjuvant chemotherapy predicts pathological response and survival in this patient group. However, no studies have evaluated the prognostic significance of metabolic parameters with respect to pathological response in lymph nodes. Change in primary tumour maximum standardised uptake value (SUVmax) is the most widely used FDG PET-CT parameter to define metabolic response, with various thresholds described. The most commonly utilised are a 35% reduction in SUVmax (MUNICON), and a 30% reduction in SUVmax (PERCIST). However, there is no consensus regarding the optimal classification; the MUNICON threshold was derived from only 40 patients and PERCIST is not tumour specific. The primary aim of this study was to evaluate the ability of FDG PET-CT to predict pathological response in the primary tumour (pTR) and lymph nodes (pNR) in patients with oesophageal adenocarcinoma undergoing neoadjuvant chemotherapy before surgery. Secondary aims were to assess the prognostic effect of metabolic and pathological response and evaluate the predictive ability of response classifications. Methods Cohort study of 75 patients with locally advanced oesophageal or oesophago-gastric junctional adenocarcinoma who underwent FDG-PET-CT before and after neoadjuvant chemotherapy, prior to surgical resection at Guy's and St Thomas’ NHS Foundation Trust, London, UK, between 2017–2020. SUV metrics related to the primary tumour and loco-regional lymph nodes were derived on pre- and post- treatment FDG PET-CT. pTR and pNR were evaluated using the Mandard classification. Patients with Mandard scores of 1–3 were classified as pathological responders and those with Mandard scores of 4–5 as non-responders. Clinicopathological characteristics were compared using the Chi-squared test. Receiver operator characteristic (ROC) analysis was performed and area under the curve (AUC) calculated to determine optimum SUVmax thresholds for metabolic response in the primary tumour (mTR) and lymph nodes (mNR). Survival curves were created using the Kaplan-Meier method, with subgroups compared using the log-rank test. Survival analysis was performed using Cox proportional hazards regression providing hazard ratios (HR) with 95% confidence intervals (CI) adjusted for age (continuous), sex (male or female), chemotherapy regimen (ECX or FLOT), cT stage (cT1–2 or cT3–4), cN stage (cN0 or cN+), tumour grade (well / moderately differentiated or poorly differentiated) and presence of signet ring cells (yes or no). Results Mean age was 63 years with the majority male (86.7%). Almost two thirds received FLOT (48/75, 64.0%) with the remainder receiving ECX chemotherapy. There was discordance between pathological response in the primary tumour and lymph nodes in several patients, including 23.1% (6/26) who demonstrated a lymph node response in the absence of a response in the primary tumour. ROC analysis demonstrated an optimum tumour SUVmax decrease of 51.2% for predicting pTR. Using a pragmatic cut-off of 50% this provided better prediction of pTR (AUC 0.714, sensitivity 73.5%, specificity 69.2%, p&amp;lt;0.001) than PERCIST (AUC 0.631, sensitivity 87.8%, specificity 38.5%, p=0.008) and MUNICON (AUC 0.659, sensitivity 85.7%, specificity 46.2%, p=0.003) criteria. ROC analysis demonstrated an optimum nodal SUVmax decrease of 32.6% for predicting pNR or pathological node negativity. Using a pragmatic 30% SUVmax cut-off and excluding metabolically negative nodes, mNR demonstrated high sensitivity but low specificity (AUC 0.749, sensitivity 92.6%, specificity 57.1%, p=0.010) for predicting pNR. pTR, mTR, pNR and mNR were independent predictive factors for overall survival on adjusted analysis (pTR responder HR 0.10 95% CI 0.03–0.34; mTR responder HR 0.17 95% CI 0.06–0.48; pNR responder HR 0.17 95% CI 0.06–0.54; mNR responder HR 0.13 95% CI 0.02–0.66). Conclusions In this study metabolic response on FDG PET-CT after neoadjuvant chemotherapy was predictive of pathological response in both the primary tumour and lymph node metastases of patients with oesophageal adenocarcinoma. Patients who had a favourable metabolic or pathological response in the primary tumour or lymph nodes had improved survival compared to non-responders. It has been suggested that commonly utilised thresholds of SUVmax are not optimal for response assessment in this patient group. This is supported by the results of the present study which suggest a reduction in SUVmax of 50% in the primary tumour was not only a better predictor of pathologic response, but also tumour recurrence and survival compared with PERCIST and MUNICON criteria. No previous studies have evaluated the ability of FDG PET-CT to predict pathologic nodal response despite recent findings suggesting it is an independent predictor of survival and evidence of a discrepancy, in some patients, who demostrate a nodal response in the absence of a response in the primary tumour. The use of FDG PET-CT in assessing response to neoadjuvant chemotherapy remains an under researched area and further evaluation is needed to establish whether it could be used to tailor individualised treatment strategies.

Utility of Circulating Free DNA Fragmentomics in the Prediction of Pathological Response after Neoadjuvant Chemoradiotherapy in Locally Advanced Rectal Cancer.

A "Watch and Wait" (W&W) approach has become an alternative to surgery for locally advanced rectal cancer (LARC) after neoadjuvant chemoradiotherapy (nCRT). Precise prediction of pathological complete response (pCR) will improve patient selection for W&W. We investigated the utility of cell-free DNA (cfDNA) fragmentomics in predicting pCR. We recruited 119 LARC patients and evaluated nCRT response by pCR status and pathological or MRI tumor regression grade (mrTRG). Plasma samples before, during, and after nCRT were applied to deep targeted-panel sequencing, with 103 patients having complete samples. cfDNA fragment and 5'-end motif profiles were used to construct elastic-net logistic regression models to predict non-pCR. Predictive performance was measured by area under the receiver operator characteristic curve (AUC), sensitivity, and specificity. In the training cohort, the model based on 5'-end motif profile plus mrTRG achieved the highest cross-validation AUC (0.92, 95% CI, 0.91-0.93). The AUC in a testing cohort was 0.96 (95% CI, 0.90-1.00). The models based on 5'-end motif profile alone or in combination with mrTRG both maintained good predictive ability for patients without detectable circulating tumor DNA (AUC 0.94, 95% CI, 0.93-0.95; AUC 0.95, 95% CI, 0.94-0.96). In an external validation cohort, the model trained with a local 5'-end motif profile obtained an AUC of 0.878 (95% CI, 0.801-0.956) in discriminating colorectal cancer from healthy subjects. The combination of a 5'-end motif profile with mrTRG has the potential to predict the response to nCRT, and therefore may improve the patient selection for a W&W approach.

Shear Wave Elastography Combined with Molecular Subtype in Early Prediction of Pathological Response to Neoadjuvant Chemotherapy in Patients with Breast Cancer: A Prospective Study

Neoadjuvant chemotherapy (NAC) is an important part of breast cancer therapy with the advantages of contributing to better survival and downstaging cancer to increase the chances of surgery or breast-conserving surgery (1,2). However, patients with chemotherapy resistance cannot benefit from NAC and may suffer from adverse reactions due to unnecessary chemotherapy. Furthermore, the timing of the operation may be delayed. Previous investigations demonstrated that the overall response rate for NAC can range from 69%‒100%, with pathological complete response being observed in 10%‒15% of patients (3).

Obesity may not be related to pathologic response in locally advanced rectal cancer following neoadjuvant chemoradiotherapy.

BackgroundThe aim of this study is to evaluate the correlation between body mass index (BMI) and body fat composition (measured with radiological fat parameters (RFP)) and pathological response after neoadjuvant chemoradiotherapy for locally advanced rectal cancer patients. The secondary aim of the study was to assess the role of BMI and RFP on major surgical complications, overall survival (OS), and disease-free survival (DFS).MethodsAll patients who underwent surgical resection following nCRT between 2005 and 2017 for mid-low rectal cancer were retrospectively collected. Visceral fat area (VFA), superficial fat area (SFA), visceral/superficial fat area ratio (V/S), perinephric fat thickness (PNF), and waist circumference (WC) were estimated by baseline CT scan. Predictors of pathologic response and postoperative complications were investigated using logistic regression analysis. The correlations between BMI and radiologic fat parameters and survival were investigated using the Kaplan–Meier method and log-rank test.ResultsOut of 144 patients included, a complete (TRG1) and major (TRG1+2) pathologic response was reported in 32 (22%) and 60 (45.5%) cases, respectively. A statistically significant correlation between BMI and all the RFP was found. At a median follow-up of 60 (35–103) months, no differences in terms of OS and DFS were found considering BMI and radiologic fat parameters. At univariable analysis, neither BMI nor radiologic fat parameters were predictors of complete or major pathologic response; nevertheless, VFA, V/S>1, and BMI were predictors of postoperative major complications.ConclusionsWe found no associations between BMI and body fat composition and pathological response to nCRT, although VFA, V/S, and BMI were predictors of major complications. BMI and RFP are not related to worse long-term OS and DFS.

Dynamic contrast-enhanced and diffusion-weighted MR imaging in early prediction of pathologic response to neoadjuvant chemotherapy in locally advanced gastric cancer.

To investigate the efficacy of diffusion-weighted imaging (DWI) and dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) for the early prediction of the pathologic response to neoadjuvant chemotherapy (NAC) in patients with locally advanced gastric cancer (LAGC). Fifty patients with LAGC who were treated with NAC followed by radical gastrectomy were enrolled. Uncontrasted and DCE-MRI were performed within 1week before NAC. According to tumor regression grading (TRG), patients were labeled as responders (TRG = 0 + 1) and non-responders (TRG = 2 + 3). Apparent diffusion coefficients (ADC) and DCE-MRI kinetics (Ktrans, Ve, and Kep) were compared between the two groups. Logistic regression analysis was performed to screen independent factors to predict the NAC efficacy. The relationship between MRI parameters and TRG was studied by Spearman's correlation analysis. Receiver-operating characteristic curve analyses were applied to evaluate the efficacy. ADC, Ktrans, and Kep values were higher in responders than in non-responders (p < 0.05) and correlated with TRG (p < 0.05). The ADC and Kep values were independent markers for predicting TRG. The area under the curve, sensitivities, specificities of ADC, Ktrans, Kep, and ADC + Kep were 0.813, 0.699, 0.709, 0.886;73.64%, 65.54%, 63.21%, 70.37%; 86.47%, 54.97%, 79.47%, 95.65%; respectively. ADC + Kep demonstrated a higher efficacy than Ktrans and Kep (p = 0.012, 0.011), but without improvement compared with ADC (p > 0.05). Both DWI and DCE-MRI can effectively predict the pathologic response to NAC in LAGC. A combination of ADC and Kep increased the efficacy, and ADC is the most valuable imaging parameter.

Machine Learning Models and Multiparametric Magnetic Resonance Imaging for the Prediction of Pathologic Response to Neoadjuvant Chemotherapy in Breast Cancer

Simple SummaryAchieving pathological complete response (pCR) to neoadjuvant chemotherapy (NAC) in breast cancer (BC) is crucial, as pCR is a surrogate marker for survival. However, only 10–30% of patients achieve it. It is therefore essential to develop tools that enable to accurately predict response. Recently, different studies have demonstrated the feasibility of applying machine learning approaches to non-invasively predict pCR before NAC administration from magnetic resonance imaging (MRI) data. Some of those models are based on radiomics, an emerging field that allows the automated extraction of clinically relevant information from radiologic images. However, the research is still at an early stage and further data are needed. Here, we determine whether the combination of imaging data (radiomic features and perfusion/diffusion imaging biomarkers) extracted from multiparametric MRIs and clinical variables can improve pCR prediction to NAC compared to models only including imaging or clinical data, potentially avoiding unnecessary treatment and delays to surgery.Background: Most breast cancer (BC) patients fail to achieve pathological complete response (pCR) after neoadjuvant chemotherapy (NAC). The aim of this study was to evaluate whether imaging features (perfusion/diffusion imaging biomarkers + radiomic features) extracted from pre-treatment multiparametric (mp)MRIs were able to predict, alone or in combination with clinical data, pCR to NAC. Methods: Patients with stage II-III BC receiving NAC and undergoing breast mpMRI were retrospectively evaluated. Imaging features were extracted from mpMRIs performed before NAC. Three different machine learning models based on imaging features, clinical data or imaging features + clinical data were trained to predict pCR. Confusion matrices and performance metrics were obtained to assess model performance. Statistical analyses were conducted to evaluate differences between responders and non-responders. Results: Fifty-eight patients (median [range] age, 52 [45–58] years) were included, of whom 12 showed pCR. The combined model improved pCR prediction compared to clinical and imaging models, yielding 91.5% of accuracy with no false positive cases and only 17% false negative results. Changes in different parameters between responders and non-responders suggested a possible increase in vascularity and reduced tumour heterogeneity in patients with pCR, with the percentile 25th of time-to-peak (TTP), a classical perfusion parameter, being able to discriminate both groups in a 75% of the cases. Conclusions: A combination of mpMRI-derived imaging features and clinical variables was able to successfully predict pCR to NAC. Specific patient profiles according to tumour vascularity and heterogeneity might explain pCR differences, where TTP could emerge as a putative surrogate marker for pCR.

Can Clinical Response Predict Pathologic Response Following Neoadjuvant Chemoradiation for Esophageal Cancer?

ObjectivesApproximately 20–40% of patients with locally advanced esophageal cancer will achieve a pathologic complete response (ypCR) following neoadjuvant chemoradiotherapy (nCRT). Predicting ypCR based on a clinical complete response (ycCR) has been a challenge. This study assessed the correlation between ycCR and ypCR, as determined from esophagectomy specimens. MethodsPatients undergoing esophagectomy following nCRT at three major institutions between 2005 and 2018 were reviewed. Restaging, including PET/CT, endoscopy with biopsy, and esophageal ultrasound (EUS), was performed to determine ycCR. ResultsSix hundred sixty patients were included, with 93.3% with esophageal adenocarcinoma histology. Six hundred fifty-eight of these patients underwent PET, 304 EUS, and 584 underwent a biopsy. Following nCRT, 148 (22.4%) were found to have a ypCR. Only 12/32 (37.5%) determined to have a ycCR were found to have a ypCR, while 136/628 (21.6%) with a non-ycCR were found to have a ypCR (p 0.075). Individual modality PPV was 28% for PET, 54% for EUS, and 26% for biopsy. When PET was combined with EUS, 168 reports were concordant and the PPV of ypCR was 50%, though the number of patients was low (1/2). With all 3 re-staging modalities combined, the PPV and NPV both rose to 100%. ConclusionsCurrent restaging tools cannot reliably predict ypCR after nCRT. While multimodal restaging appears to be a more accurate predictor of ypCR than any testing modality alone, patients cannot reliably be advised to avoid an esophagectomy on the assumption that ycCR predicts ypCR at this time.

Prediction of pathological response to preoperative chemotherapy for pancreatic ductal adenocarcinoma using 2-[18F]-fluoro-2-deoxy-d-glucose positron-emission tomography

To determine whether the pathological response to preoperative chemotherapy for pancreatic ductal adenocarcinoma (PDAC) can be predicted using 2-[18F]-fluoro-2-deoxy-d-glucose positron-emission tomography (F-18 FDG-PET). Twenty-eight patients with PDAC who underwent only neoadjuvant chemotherapy (NAC) before surgery were enrolled in the study. All patients had F-18 FDG-PET examinations before NAC. The resected specimen was pathologically evaluated according to the Classification of Pancreatic Carcinoma (7th edn). Patients were categorised into a non-response group and a response group based on the pathological findings. The non-response group (Grades 1a and 1b) showed ≤50% necrosis in the specimen, while the specimens of the response group (Grades 2-3) showed >50% necrosis. The maximum standardised uptake values (SUVmax) of the tumours on F-18 FDG-PET were measured. The mean values of SUVmax were compared between the two groups. The diagnostic performance of SUVmax in distinguishing the two groups was also evaluated using receiver operating characteristic analysis. The mean SUVmax of the response group was higher than that of the non-response group (9.00±1.78 versus 4.26±2.35; p<0.001). The optimal cut-off value of SUVmax was 9.28 for distinguishing the two groups. The sensitivity, specificity, and accuracy for the prediction in the response group were 80%, 95.7%, and 92.9%, respectively. SUVmax on F-18 FDG-PET may be useful as a biomarker to predict the pathological response to NAC in patients with PDAC.

Contrast-free MRI quantitative parameters for early prediction of pathological response to neoadjuvant chemotherapy in breast cancer.

To assess early changes in synthetic relaxometry after neoadjuvant chemotherapy (NAC) for breast cancer and establish a model with contrast-free quantitative parameters for early prediction of pathological response. From March 2019 to January 2021, breast MRI were performed for a primary cohort of women with breast cancer before (n = 102) and after the first (n = 93) and second (n = 90) cycle of NAC. Tumor size, synthetic relaxometry (T1/T2 relaxation time [T1/T2], proton density), and ADC were obtained, and the changes after treatment were calculated. Prediction models were established by multivariate logistic regression; evaluated with discrimination, calibration, and clinical application; and compared with Delong tests, net reclassification (NRI), and integrated discrimination index (IDI). External validation was performed from February to June 2021 with an independent cohort of 35 patients. In the primary cohort, all parameters changed after early treatment. Synthetic relaxometry decreased to a greater degree in major histologic responders (MHR, Miller-Payne G4-5) compared with non-MHR (Miller-Payne G1-3). A model combining ADC after treatment, changes in T1 and tumor size, and cancer subtype achieved the highest AUC after the first (primary/validation cohort, 0.83/0.82) and second cycles (primary/validation cohort, 0.85/0.84). No difference of AUC (p ≥ 0.27), NRI (p ≥ 0.31), and IDI (p ≥ 0.32) was found between models with different cycles and size-measured sequences. Model calibration and decision curves demonstrated a good fitness and clinical benefit, respectively. Early reduction in synthetic relaxometry indicated pathological response to NAC. Contrast-free T1 and ADC combined with size and cancer subtype predicted effectively pathological response after one NAC cycle. • Synthetic MRI relaxometry changed after early neoadjuvant chemotherapy, which demonstrated pathological response for mass-like breast cancers. • Contrast-free quantitative parameters including T1 relaxation time and apparent diffusion coefficient, combined with tumor size and cancer subtype, stratified major histologic responders. • A contrast-free model predicted an early pathological response after the first treatment cycle of neoadjuvant chemotherapy.