Radiological Response Assessment Research Articles

Safety and Tolerability of Interval Compressed Chemotherapy Schedule in Children Receiving Treatment for Ewing Sarcoma: A Real-world Experience From India.

Chemotherapy with alternating cycles of vincristine-doxorubicin-cyclophosphamide and ifosfamide-etoposide, along with primary tumor treatment with surgery or radiotherapy or both, constitute the usual treatment of Ewing sarcoma. The AEWS0031 study demonstrated survival benefits after interval-compressed chemotherapy without significant toxicity. The aim of this study was to assess the tolerability of dose-intensified chemotherapy in developing countries like India. This was a retrospective analysis of children younger than 18 years of age with newly diagnosed Ewing sarcoma who came for treatment from December 2017 to December 2022. Children received vincristine (2mg/m2), doxorubicin (75mg/m2), and cyclophosphamide (1.2g/m2) alternating with ifosfamide (9g/m2), etoposide (500mg/m2) for 17 cycles, with filgrastim (5µg/kg; maximum 300µg) between cycles. Primary tumor treatment was provided with surgery or radiotherapy or both. Local treatment was given between weeks 12 to 16. Toxicity was assessed using the Common Terminology Criteria for Adverse Events (CTCAE) Version 3.0. Radiologic response assessment was carried out with restaging CT or MRI scans after 6 to 8 cycles of chemotherapy in nonmetastatic and metastatic settings, respectively. Thirty-one children were enrolled. Twenty-three children received all 17 cycles of chemotherapy. The median cycle interval was 18 days and 41% of children received chemotherapy at the 2-week interval. Grade 4 febrile neutropenia was observed in 32% of cycles but no treatment-related mortality was reported. Anemia and thrombocytopenia requiring transfusion support were recorded in 28 (5.6%) and 69 cycles (13.9%) of chemotherapy, respectively. There were 2 events of grade 4 cardiac toxicities in the form of cardiomyopathy and arrhythmia requiring intensive care management. After surgery, good necrosis was achieved in 61% of cases. Three children had a relapse with an event-free survival (EFS) of 87%. Intensified chemotherapy administered every 2 weeks intervals in Ewing sarcoma, is tolerable with adequate supportive care in resource-constrained settings.

Computed Tomography-Based Radiomics with Machine Learning Outperforms Radiologist Assessment in Estimating Colorectal Liver Metastases Pathologic Response After Chemotherapy.

This study was designed to assess computed tomography (CT)-based radiomics of colorectal liver metastases (CRLM), extracted from posttreatment scans in estimating pathologic treatment response to neoadjuvant therapy, and to compare treatment response estimates between CT-based radiomics and radiological response assessment by using RECIST 1.1 and CT morphologic criteria. Patients who underwent resection for CRLM from January 2003-December 2012 at a single institution were included. Patients who did not receive preoperative systemic chemotherapy, or without adequate imaging, were excluded. Imaging characteristics were evaluated based on RECIST 1.1 and CT morphologic criteria. A machine-learning model was designed with radiomic features extracted from manually segmented posttreatment CT tumoral and peritumoral regions to identify pathologic responders (≥ 50% response) versus nonresponders. Statistical analysis was performed at the tumor level. Eighty-five patients (median age, 62 years; 55 women) with 95 tumors were included. None of the subjectively evaluated imaging characteristics were associated with pathologic response (p > 0.05). Inter-reader agreement was substantial for RECIST categorical response assessment (K = 0.70) and moderate for CT morphological group response (K = 0.50). In the validation cohort, the machine learning model built with radiomic features obtained an area under the curve (AUC) of 0.87 and outperformed subjective RECIST assessment (AUC = 0.53, p = 0.01) and morphologic assessment (AUC = 0.56, p = 0.02). Radiologist assessment of oligometastatic CRLM after neoadjuvant therapy using RECIST 1.1 and CT morphologic criteria was not associated with pathologic response. In contrast, a machine-learning model based on radiomic features extracted from tumoral and peritumoral regions had high diagnostic performance in assessing responders versus nonresponders.

Immunotherapy response and resistance in patients with advanced uveal melanoma: a retrospective cohort study

Metastatic uveal melanoma (mUM) is associated with poor prognosis. Ipilimumab/nivolumab has shown antitumor efficacy in phase II studies. Tebentafusp resulted in longer overall survival (OS) compared to investigator`s choice in a phase III study. We sought to describe the radiological response patterns of mUM patients treated with immunotherapy. Patients with mUM treated with ipilimumab/nivolumab and tebentafusp between July 2018 and December 2022, with available radiological assessment per RECISTv1.1 and/or imPERCIST5, were retrospectively identified and included. Progression-free survival (PFS) and OS rates, liver-specific response and pathological assessment in available liver biopsies were evaluated. In the ipilimumab/nivolumab group, median PFS (mPFS) was 2.9 months (95% CI 2.2–28.6) and mOS 28.9 months (95% CI 12.7-NR). Complete (CMR) and partial (PMR) metabolic response per imPERCIST5, and partial response (PR) per RECISTv1.1 were associated with longer PFS and OS by trend, compared to morphologically and metabolically stable or progressive disease. In the tebentafusp group, mPFS was 2.7 months (95% CI 2.2–3) and mOS 18.6 months (95% CI 11.5-NR). PMR and PR were associated with longer PFS by trend. In both treatments, the overall treatment response was associated with the radiological response at the liver site. In available liver tumor biopsies, differences in pathological and radiological responses were noted. ImPERCIST5 and RECIST v1.1 are valuable tools in the radiological response assessment, but both methods display limitations. Accurate biomarkers to stratify patients at risk for disease progression and future translational studies to investigate mechanisms of response and resistance are required.

Automated Lugano Metabolic Response Assessment in 18F-Fluorodeoxyglucose-Avid Non-Hodgkin Lymphoma With Deep Learning on 18F-Fluorodeoxyglucose-Positron Emission Tomography.

Artificial intelligence can reduce the time used by physicians on radiological assessments. For 18F-fluorodeoxyglucose-avid lymphomas, obtaining complete metabolic response (CMR) by end of treatment is prognostic. Here, we present a deep learning-based algorithm for fully automated treatment response assessments according to the Lugano 2014 classification. The proposed four-stage method, trained on a multicountry clinical trial (ClinicalTrials.gov identifier: NCT01287741) and tested in three independent multicenter and multicountry test sets on different non-Hodgkin lymphoma subtypes and different lines of treatment (ClinicalTrials.gov identifiers NCT02257567, NCT02500407; 20% holdout in ClinicalTrials.gov identifier NCT01287741), outputs the detected lesions at baseline and follow-up to enable focused radiologist review. The method's response assessment achieved high agreement with the adjudicated radiologic responses (eg, agreement for overall response assessment of 93%, 87%, and 85% in ClinicalTrials.gov identifiers NCT01287741, NCT02500407, and NCT02257567, respectively) similar to inter-radiologist agreement and was strongly prognostic of outcomes with a trend toward higher accuracy for death risk than adjudicated radiologic responses (hazard ratio for end of treatment by-model CMR of 0.123, 0.054, and 0.205 in ClinicalTrials.gov identifiers NCT01287741, NCT02500407, and NCT02257567, compared with, respectively, 0.226, 0.292, and 0.272 for CMR by the adjudicated responses). Furthermore, a radiologist review of the algorithm's assessments was conducted. The radiologist median review time was 1.38 minutes/assessment, and no statistically significant differences were observed in the level of agreement of the radiologist with the model's response compared with the level of agreement of the radiologist with the adjudicated responses. These results suggest that the proposed method can be incorporated into radiologic response assessment workflows in cancer imaging for significant time savings and with performance similar to trained medical experts.

Preliminary efficacy and safety of camrelizumab combined with cetuximab and chemotherapy in recurrent or metastatic head and neck squamous cell carcinoma (R/M HNSCC): A phase II clinical trial.

e18011 Background: Cetuximab and PD-1 inhibitors are widely used first-line therapy for R/M HNSCC and may provide potential synergy in R/M HNSCC. We conducted an open-label, single-arm, Simon's two-stage, phase II study of camrelizumab with cetuximab and cisplatin-based chemotherapy as first-line treatment in R/M HNSCC (NCT05673577). We report the results of stage 1 of this study. Methods: Eligible patients with R/M HNSCC not amenable to curative treatment were enrolled. Patients were treated with camrelizumab 200mg intravenously Q3W, cetuximab 400mg/m2 loading dose followed by 250mg/m2 weekly (21-day cycle), cisplatin 75mg/m2 Q3W, for up to 6 cycles, and nab-paclitaxel 125mg/m2 on d1, d8 (21-day cycle), for up to 6 cycles. Maintenance therapy with camrelizumab 200mg intravenously Q2W, cetuximab 500mg/m2 Q2W were given until intolerable toxicity or disease progression. Simon-minimax design was applied, if 9 of the 21 patients achieve PR or CR in the first stage, another 19 patients will be included. The primary endpoint is ORR. Secondary endpoints include PFS, OS, safety, etc, andmolecular biomarkers will be tested as exploratory endpoints. Results: A total of 21 patients were enrolled in the stage 1, with a median age of 65y (range 34-72y), M:F 19:2,ECOG (0:1) 1:20, and primary sites of oral cavity (n=6), oropharynx (n=3), hypopharynx (n=2), larynx (n=8), and unknown (n=2). By the cutoff date of Jan 10 2024, 20 patients were followed up (1 withdrew because of refusing and discontinuing treatment after SD). In total, 2 (9.5%) CR, 18 (85.7%) PR and 1 (4.8%) SD were observed as best response for an ORR of 95.2% and a DCR of 100%, meeting pre-planned criteria for trial continuation. Meanwhile, 18 patients achieved PR and 1 patient achieved SD after receiving the first two cycles of treatment (2 patients didn't complete accurate radiologic assessment of tumor response after 2 cycles of treatment and were confirmed PR after 4 cycles). Out of 21, 13 patients completed planned cycles of treatment and started maintenance therapy, 5 patients were still under planned cycles. The only one case of progressive disease occurred during the maintenance period. Grade 3 or above treatment-related adverse events were observed in 47.6% patients. One patient discontinued camrelizumab due to pneumonia, and one patient discontinued cetuximab due to grade 3 rash. Three patients reported fatal AEs, which was caused by tumor hemorrhage in the re-irradiated site, severe infectious pneumonia and choking due to swallowing, respectively. Conclusions: Camrelizumab combined with cetuximab and cisplatin-based chemotherapy induced a high ORR with prompt tumor shrinkage and was well tolerated in the scenario of first-line R/M HNSCC. These results meet protocol specifications for trial continuation. Final results will include long-term and stage 2 results. Clinical trial information: NCT05673577 .

A phase 2, single arm, European multi-center trial evaluating the efficacy of afatinib as first line or later line treatment in advanced chordoma.

11517 Background: EGFR and Her2 overexpression in chordoma is well known, and chordoma cell-lines and mouse models were proven to be sensitive to EGFR inhibitor afatinib. This phase 2, single arm, European multi-center trial was designed to evaluate the efficacy of afatinib as first- or later-line tyrosine kinase inhibitor (TKI) treatment in advanced chordoma. Here we report efficacy and safety data. Methods: Eligible patients (pts) had locally advanced or metastatic, pathologically proven, EGFR expressing chordoma, not amenable for local therapies, with confirmed measurable and progressive lesions according to RECIST1.1. Pts were treated with afatinib 40mg/day in a 4 week cycle until disease progression (PD), unacceptable toxicity or withdrawal. Radiological tumor response assessment was performed every 3 cycles. The primary endpoint was response defined as progression free survival (PFS) rate ≥12 months (mos) for first-line cohort 1 and ≥ 9 mos for further-line treatment cohort 2, and change from baseline in EORTC QLQ- C30 and Brief pain inventory (BPI) questionnaires. Pts were enrolled using a Simon’s two-stage design, enrolling 13 patients in stage one, ≥3 patients with a PFS ≥ 12 or 9 mos were needed to enter stage two, where 43 pts total were to be enrolled, and ≥13 free from progression at 9 or 12 mos were required to meet the primary endpoint for success. Results: From Jun 2018 to Oct 2022, 47 pts were included. Four were ineligible for efficacy analysis (1 withdrawal, 3 stopped due to toxicity before first radiological evaluation). 34 entered cohort 1, 13 cohort 2. 31 (66.0%) were men, median age was 53 (range 28-85) years. 16 (34.0%) pts received prior systemic therapy (3 chemotherapy, 13 TKI, 2 immunotherapy, 2 other). The PFS rate at 12 mos was 40.0% in cohort 1 (12/30 pts), and 38.5% in cohort 2 (5/13 pts). Overall median PFS was 8.6 mos (95% CI 5.6-13.6); 9.1 mos (95% CI 5.8-16.6) in cohort 1 and 7.1 mos (95% CI 2.8-16.5) in cohort 2. Two pts remained on treatment at time of analysis. Best objective response by RECIST 1.1 was partial response in 4 pts (9.3%), stable disease in 33 (76.7%), PD in 5 pts (11.6%) and 1 (2.3%) unknown due to clinical progression. Median follow-up time was 23.7 mos (interquartile range 11.0-21.2). 16/47 pts (30.4%) experienced grade ≥3 adverse events (AEs). No grade 5 AEs related to afatinib were reported. Most common grade 3-4 afatinib-related AEs were skin toxicity (10.6% of pts), diarrhea (10.6%), mucositis (6.4%), hypertension (4.3%). Dose reduction was needed in 20/47 (42.6%) pts, and at least one dose interruption was needed in 31/47 (66.0%) pts. Conclusions: With a PFS rate at 12 mos of 40.0% in the first-line cohort and 38.5% at 9 mos in the further-line cohort, this phase 2 study met the PFS endpoint. QoLQ data will be provided at the conference . Partial response by RECIST was seen in 4/43 pts. Toxicity and dose reductions were relevant but manageable in most pts. Clinical trial information: NCT03083678 .

217P Less frequent radiological response assessment to reduce costs from 177-LuDOTATE therapy for patients with advanced neuroendocrine tumors

217P Less frequent radiological response assessment to reduce costs from 177-LuDOTATE therapy for patients with advanced neuroendocrine tumors

Abstract 2522: Predictive value of tumor mutational burden (TMB) in patients with metastatic microsatellite-stable (MSS) colorectal cancer (CRC) given first-line oxaliplatin-based chemotherapy and immune checkpoint blockade (ICB)

Abstract Purpose: Most CRC patients harbor primarily non-immunogenic MSS tumor. The randomized METIMMOX trial (NCT03388190) examined if patients with previously untreated, unresectable abdominal metastases from MSS-CRC may achieve therapeutic efficacy from short-course oxaliplatin-based chemotherapy (FLOX) and sequential ICB (nivolumab) repeatedly. Half of study patients assigned to this experimental treatment had significantly extended progression-free survival (PFS) while the other half had shortened PFS, compared to the control-group patients given standard FLOX chemotherapy with median PFS 9.3 months. Here we explored if somatic mutations (mut) unveiled by next-generation sequencing might provide insights into responsiveness to the METIMMOX regimen. Procedures: Patients had MSS-CRC with infradiaphragmatic metastases deemed unresectable and were eligible for first-line oxaliplatin-based therapy. They were randomly assigned to the control group of FLOX (oxaliplatin, 5-fluorouracil, folinic acid) Q2W or the experimental group of alternating 2 cycles each of FLOX Q2W and nivolumab Q2W, with prespecified break periods. Radiologic response assessment was done every 8 weeks with PFS as the primary endpoint. Diagnostic tumor biopsies and baseline metastasis biopsies were sequenced with the TruSight Oncology 500 assay. TMB was calculated using only coding, non-synonymous single-nucleotide variants and insertions/deletions with variant allele frequency ≥5% and sequencing depth ≥50 ×, applying the effective panel size as the megabases (Mb) denominator. Results: Sequencing data were acquired from 20 experimental-group patients with median PFS 7.9 months (minimum 2.0, maximum 41.6). Median TMB was 5.1 mut/Mb (minimum 0.8, maximum 11.8), concordant in patient-paired primary tumor and metastasis specimens. The TMB was associated with treatment outcome; the higher TMB, the better PFS with hazard ratio 0.83 (95% confidence interval (CI) 0.70-0.98; p = 0.026, Cox regression) for a PFS event with increasing TMB. TMB cut-off as low as 5.0 mut/Mb distinguished between patient groups (p = 0.007, log-rank test) with median PFS 34.9 months (95% CI 10.6-59.2; TMB>5, n = 10) and median PFS 4.6 months (95% CI 2.7-6.5; TMB≤5, n = 10). In the TMB>5 group, 8/10 were KRAS- or BRAF-mutant cases. In the TMB≤5 group, 9/10 metastasis specimens were RAS/BRAF-wildtype cases. Conclusions: TMB >5 mut/Mb was associated with tumor KRAS or BRAF driver mutation and remarkably long PFS in first-line treatment of patients with abdominal metastases from MSS-CRC with alternating short-course oxaliplatin-based chemotherapy and ICB. Citation Format: Anne Hansen Ree, Paula A. Bousquet, Hilde L. Nilsen, Torben Lüders, Shixiong Wang, Tina Visnovska, Diana L. Bordin, Eirik Høye, Hanne M. Hamre, Christian Kersten, Eva Hofsli, Marianne G. Guren, Halfdan Sorbye, Kjersti Flatmark, Sebastian Meltzer. Predictive value of tumor mutational burden (TMB) in patients with metastatic microsatellite-stable (MSS) colorectal cancer (CRC) given first-line oxaliplatin-based chemotherapy and immune checkpoint blockade (ICB) [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 2522.

Tumor growth rate to assess therapy response to immune-based combinations for metastatic renal cell carcinoma

ABSTRACT Background Radiological response assessment is becoming challenging with novel immune-based combinations for metastatic renal cell carcinoma (mRCC). RECIST criteria appear not exhaustively adequate to capture the kinetics of treatment response, which is better reflected by tumor growth rate (TGR). We explored TGR changes during first-line treatments and its association with clinical outcomes in mRCC. Research design and methods We retrospectively evaluated TGR in untreated patients undergoing pembrolizumab/axitinib (P/A) or tyrosine-kinase inhibitors (TKI). TGR was calculated at the first (TGR1, after 3 months) and the second (TGR2, after 6 months) evaluation, thus assessing the TGR2-TGR1 difference. Results Thirty-three patients were included (P/A n = 15, TKIs n = 18). Volumes firstly decreased more rapidly with TKIs, and then more slowly. Volumes initially remained stable with P/A, quickly decreasing until the second evaluation. TGR1 was related to progression-free survival (PFS; p = 0.023) and overall survival (p = 0.046) with P/A. TGR2 was correlated with PFS in all patients (p = 0.025). Patients with higher velocity volume reduction appeared to have improved survival benefits than patients with lower velocity considering both treatments, but especially with P/A. Conclusion Combining immunotherapy with TKIs has an important role in enhancing the rapidity of tumor shrinkage. A rapid disease volume reduction correlates with better OS and PFS.

Early radiologic and metabolic tumour response assessment during combined chemo-radiotherapy for locally advanced NSCLC

BackgroundThe role of early treatment response for patients with locally advanced non-small cell lung cancer (LA-NSCLC) treated with concurrent chemo-radiotherapy (cCRT) is unclear. The study aims to investigate the predictive value of response to induction chemotherapy (iCX) and the correlation with pattern of failure (PoF). Materials and methodsPatients with LA-NSCLC treated with cCRT were included for analyses (n = 276). Target delineations were registered from radiotherapy planning PET/CT to diagnostic PET/CT, in between which patients received iCX. Volume, sphericity, and SUVpeak were extracted from each scan. First site of failure was categorised as loco-regional (LR), distant (DM), or simultaneous LR+M (LR+M). Fine and Gray models for PoF were performed: a baseline model (including performance status (PS), stage, and histology), an image model for squamous cell carcinoma (SCC), and an image model for non-SCC. Parameters included PS, volume (VOL) of tumour, VOL of lymph nodes, ΔVOL, sphericity, SUVpeak, ΔSUVpeak, and oligometastatic disease. ResultsMedian follow-up was 7.6 years. SCC had higher sub-distribution hazard ratio (sHR) for LRF (sHR = 2.771 [1.577:4.87], p < 0.01) and decreased sHR for DM (sHR = 0.247 [0.125:0.485], p < 0.01). For both image models, high diagnostic SUVpeak increased risk of LRF (sHR = 1.059 [1.05:1.106], p < 0.01 for SCC, sHR = 1.12 [1.03:1.21], p < 0.01 for non-SCC). Patients with SCC and less decrease in VOL had higher sHR for DM (sHR = 1.025[1.001:1.048] pr. % increase, p = 0.038). ConclusionPoor response in disease volume was correlated with higher sHR of DM for SCC, no other clear correlation of response and PoF was observed. Histology significantly correlated with PoF with SCC prone to LRF and non-SCC prone to DM as first site of failure. High SUVpeak at diagnosis increased the risk of LRF for both histologies.

Enrichment of circulating helper-T (Th) and unswitched memory-B (NSwM-B) cell populations and responsiveness to immune checkpoint blockade (ICB) in microsatellite-stable (MSS)/mismatch repair-proficient (pMMR) colorectal cancer (CRC).

150 Background: The anti-tumor activity of ICB in MSS/pMMR-CRC is limited. However, recent studies have highlighted the possibility of inducing ICB responsiveness by oxaliplatin-based chemotherapy. The understanding of underlying mechanisms and identification of rational biomarkers are major focus areas in immuno-oncology. Methods: Patients with previously untreated, unresectable abdominal metastases from MSS/pMMR-CRC were randomly assigned to the oxaliplatin-based standard Nordic FLOX regimen (control arm) or 2 FLOX cycles followed by 2 ICB (nivolumab) cycles in a repeat sequential schedule (experimental arm). Radiologic response assessment was done every 8 weeks, and the study arms reached identical primary endpoint – median progression-free survival (PFS) 9.3 months. Experimental-arm patients at one study center had peripheral blood mononuclear cells (PBMC) prepared at study entry (V1), after the initial 2 FLOX cycles (V3), and after the succeeding 2 nivolumab cycles (V5). The PBMC were incubated with 19 cell-surface antibodies and analyzed by mass cytometry. Patients were categorized in PFS &gt;9.3 months (improved; n = 7) or PFS &lt;9.3 months (shortened; n = 6). Live cell clusters were identified by hierarchically applying automated 1D gates to the available lineage markers. After gating, all cells were assigned to soft clusters according to phenotypes. The cluster sizes were used to compute the differential abundance of immune-cell populations and a Peacock test assessed multivariate changes between the improved and shortened PFS groups. Results: Circulating CD4+ (Th) cells and CD27+IgD+ (NSwM-B) cells were concurrently enriched in the improved PFS group. After the initial 2 FLOX cycles (V3), median increase (from V1) was 2.89% (of lymphocyte count) for Th cells and 0.54% (of leukocyte count) for NSwM-B cells (p &lt; 0.001). Continued responses occurred over the succeeding 2 nivolumab cycles (until V5) with median increase of 1.19% for Th cells and 0.19% for NSwM-B cells (p &lt; 0.001). The association between the immune-cell alterations and PFS was validated by categorizing the cases in above or below the median increase in Th and NSwM-B cell populations from V1 to V5. The high-increase group achieved median PFS 15.0 months (95% CI, 10.8-19.2) compared to 4.0 months (95% CI, 1.9-22.1) for the low-increase group (log-rank; p = 0.002). Conclusions: This unsupervised analysis of high-dimensional PBMC data from a limited case number revealed that circulating Th and NSwM-B cell populations may mediate ICB responsiveness invoked by short-course oxaliplatin-based chemotherapy in patients with abdominal metastases from MSS/pMMR-CRC. Owing to weaknesses of such a post-hoc analysis, further studies are needed to confirm the generalizability of these findings. Clinical trial information: NCT003388190 .

Deep learning-based multimodal segmentation of oropharyngeal squamous cell carcinoma on CT and MRI using self-configuring nnU-Net.

To evaluate deep learning-based segmentation models for oropharyngeal squamous cell carcinoma (OPSCC) using CT and MRI with nnU-Net. This single-center retrospective study included 91 patients with OPSCC. The patients were grouped into the development (n = 56), test 1 (n = 13), and test 2 (n = 22) cohorts. In the development cohort, OPSCC was manually segmented on CT, MR, and co-registered CT-MR images, which served as the ground truth. The multimodal and multichannel input images were then trained using a self-configuring nnU-Net. For evaluation metrics, dice similarity coefficient (DSC) and mean Hausdorff distance (HD) were calculated for test cohorts. Pearson's correlation and Bland-Altman analyses were performed between ground truth and prediction volumes. Intraclass correlation coefficients (ICCs) of radiomic features were calculated for reproducibility assessment. All models achieved robust segmentation performances with DSC of 0.64 ± 0.33 (CT), 0.67 ± 0.27 (MR), and 0.65 ± 0.29 (CT-MR) in test cohort 1 and 0.57 ± 0.31 (CT), 0.77 ± 0.08 (MR), and 0.73 ± 0.18 (CT-MR) in test cohort 2. No significant differences were found in DSC among the models. HD of CT-MR (1.57 ± 1.06mm) and MR models (1.36 ± 0.61mm) were significantly lower than that of the CT model (3.48 ± 5.0mm) (p = 0.037 and p = 0.014, respectively). The correlation coefficients between the ground truth and prediction volumes for CT, MR, and CT-MR models were 0.88, 0.93, and 0.9, respectively. MR models demonstrated excellent mean ICCs of radiomic features (0.91-0.93). The self-configuring nnU-Net demonstrated reliable and accurate segmentation of OPSCC on CT and MRI. The multimodal CT-MR model showed promising results for the simultaneous segmentation on CT and MRI. Deep learning-based automatic detection and segmentation of oropharyngeal squamous cell carcinoma on pre-treatment CT and MRI would facilitate radiologic response assessment and radiotherapy planning. • The nnU-Net framework produced a reliable and accurate segmentation of OPSCC on CT and MRI. • MR and CT-MR models showed higher DSC and lower Hausdorff distance than the CT model. • Correlation coefficients between the ground truth and predicted segmentation volumes were high in all the three models.

A European Society of Oncologic Imaging (ESOI) survey on the radiological assessment of response to oncologic treatments in clinical practice

ObjectivesTo present the results of a survey on the assessment of treatment response with imaging in oncologic patient, in routine clinical practice. The survey was promoted by the European Society of Oncologic Imaging to gather information for the development of reporting models and recommendations.MethodsThe survey was launched on the European Society of Oncologic Imaging website and was available for 3 weeks. It consisted of 5 sections, including 24 questions related to the following topics: demographic and professional information, methods for lesion measurement, how to deal with diminutive lesions, how to report baseline and follow-up examinations, which previous studies should be used for comparison, and role of RECIST 1.1 criteria in the daily clinical practice.ResultsA total of 286 responses were received. Most responders followed the RECIST 1.1 recommendations for the measurement of target lesions and lymph nodes and for the assessment of tumor response. To assess response, 48.6% used previous and/or best response study in addition to baseline, 25.2% included the evaluation of all main time points, and 35% used as the reference only the previous study. A considerable number of responders used RECIST 1.1 criteria in daily clinical practice (41.6%) or thought that they should be always applied (60.8%).ConclusionSince standardized criteria are mainly a prerogative of clinical trials, in daily routine, reporting strategies are left to radiologists and oncologists, which may issue local and diversified recommendations. The survey emphasizes the need for more generally applicable rules for response assessment in clinical practice.Critical relevance statementCompared to clinical trials which use specific criteria to evaluate response to oncological treatments, the free narrative report usually adopted in daily clinical practice may lack clarity and useful information, and therefore, more structured approaches are needed.Key points· Most radiologists consider standardized reporting strategies essential for an objective assessment of tumor response in clinical practice.· Radiologists increasingly rely on RECIST 1.1 in their daily clinical practice.· Treatment response evaluation should require a complete analysis of all imaging time points and not only of the last.Graphical

A Comparative Study Evaluating the Quality of Life and Survival Outcomes in Patients Receiving Chemotherapy Versus Oral Tyrosine Kinase Inhibitor in the Third Line and Beyond Setting for Advanced NSCLC

IntroductionThe outcomes in advanced NSCLC have improved owing to the availability of more effective systemic and improved supportive care. This has increased the number of patients who seek treatment in the third line and beyond setting. We conducted this study to compare the quality of life (QoL), toxicity, and outcomes in patients receiving chemotherapy and EGFR tyrosine kinase inhibitors (TKIs) in this setting. MethodsIn this phase 3, randomized, open-label study, patients with stage III or IV NSCLC with disease progression on at least two prior lines of chemotherapy, with a life expectancy of at least 3 months, without prior EGFR TKI exposure, and stable brain metastases (if any) were included. Patients were randomized to receive chemotherapy (gemcitabine or docetaxel or paclitaxel or vinorelbine) or an EGFR TKI (erlotinib or gefitinib). The primary end point was the change in QoL at 8 to 10 weeks; the secondary outcomes were safety and overall survival (OS). Patients underwent clinical evaluation at every visit, and toxicity was assessed as per Common Terminology Criteria for Adverse Events version 4.03. A radiological tumor response assessment was done every 8 to 12 weeks from the start of therapy. The QoL was assessed using the EORTC QLQ C30 and LC13 questionnaires. The change in QoL scores was calculated as the difference between scores at baseline and scores at 8 to 10 weeks (Δ) for each QoL domain. The Mann-Whitney U test was used to compare the mean difference (Δ) for each domain. OS and progression-free survival (PFS) were determined using the Kaplan-Meier method and Cox proportional regression analysis. ResultsA total of 246 patients were enrolled in the study, with 123 in each arm. There was a male predominance with 69.1% male patients in the chemotherapy arm and 70.7% in the EGFR TKI arm. The median age of patients in the chemotherapy arm was 54 years and 55 years in the chemotherapy and EGFR TKI arms, respectively. There was no significant difference in the change in QoL at baseline and the second visit (Δ) in both arms in all domains of EORTC QLQ C30 except cognitive function (p = 0.0045) and LC13 except alopecia (0.01249). The mean Δ Global Health Status was −28 in the chemotherapy arm and −26.8 in the EGFR TKI arm; this was not statistically significant (p = 0.973). The median follow-up was 88.1 months (95% confidence interval [CI]: 39.04–137.15). On the intention-to-treat analysis, the median PFS was 3.13 months (95% CI: 2.15–4.11) in the chemotherapy arm and 2.26 months (95% CI: 2.1–2.43) in the EGFR TKI arm, with hazard ratio at 1.074 (95% CI: 0.83–1.38) (p = 0.58). There were 120 deaths in each arm. The median OS was 7.63 months (95% CI: 5.96–9.30) in the chemotherapy arm and 7.5 months in the EGFR TKI arm (95% CI: 5.85–9.14); hazard ratio at 1.033 (95% CI: 0.80–1.33) (p = 0.805). The toxicity profile was similar in both arms except for a significantly higher incidence of fatigue (p = 0.043), peripheral neuropathy (0.000), alopecia, hypokalemia (0.037), and pedal edema (0.007) in the chemotherapy arm and dry skin (p = 0.000) and skin rash (p = 0.019) in the EGFR TKI arm. ConclusionsThere was no significant difference in most QoL scales (except cognitive function and alopecia), OS, and PFS of patients with advanced NSCLC receiving an EGFR TKI as compared with chemotherapy TKI in the third-line setting. The toxicity profile is consistent with the known toxicities of the agents.

A Phase II Study of Chidamide Plus R-GDP Salvage Chemotherapy in Relapsed/Refractory Diffuse Large B-Cell Lymphoma Ineligible for Autologous Transplantation: Efficacy and Safety Analysis

Background: A significant subset of Diffuse large B-cell lymphoma (DLBCL) patients (30%-40%) face treatment failure, necessitating effective salvage chemotherapy. We conducted a phase II study evaluating the efficacy and safety of chidamide plus R-GDP (rituximab, gemcitabine, dexamethasone, cisplatin) in relapsed/refractory (r/r) DLBCL patients. Methods: This open-label, single-arm study included 27 patients with r/r DLBCL who were ineligible for autologous stem cell transplantation. Prior to the induction monotherapy phase, chidamide was administered orally at a dose of 30 mg twice a week for one week. The combination therapy phase consisted of oral chidamide at a dose of 20 mg twice a week for two weeks, with a one-week discontinuation period, along with intravenous administration of rituximab, gemcitabine, dexamethasone, and cisplatin on a 21-day cycle. Results: A total of 27 patients received at least one dose of Chidamide plus R-GDP and were evaluated for safety. The median follow-up period was 17.0 months (range: 3.5 to 55 months). Radiological response assessment was available for 24 patients, who constituted the efficacy population. Fourteen patients completed six cycles of chidamide plus R-GDP (per-protocol population), while 10 patients discontinued treatment at the data cutoff. The baseline characteristics of enrolled patients included a median age of 67 years, predominantly older than 65 years. Approximately half of the patients exhibited elevated lactate dehydrogenase levels. All patients had received prior therapy with R-CHOP or equivalent chemoimmunotherapy. The majority of cases were classified as stage III/IV and non-germinal center B-cell-like (non-GCB) type by immunohistochemistry. Refractory or relapsed DLBCL within 1 year since the last treatment was observed in 74.0% of patients. Efficacy Outcomes: The investigator-assessed best objective response rate (ORR) was 79.2% (95% CI: 75.1% to 83.3%), with a complete response (CR) rate of 45.8% (95% CI: 41.6% to 49.9%) and a partial response (PR) rate of 33.3% (95% CI: 29.3% to 37.4%). The median time to response was 1.2 months, and the median time to CR or CRu (unconfirmed complete response) was 3.9 months. The best disease control rate was 87.5%. Among the patients who completed six cycles of chidamide plus R-GDP, the best ORR was 100%, with 71.4% achieving CR and 28.6% PR. Kaplan-Meier plots reveal that the median investigator-assessed progression-free survival (PFS) for the efficacy population was 5.9 (95% CI, 3.1 to 12.4) months. The median overall survival was 48.3 (95% CI, 13.1 to not reach (NR)) months. Safety Outcomes: A total of 13 patients (48.1%) required treatment interruption, primarily due to adverse events (22.2%) and disease progression (18.5%). Among patients who completed six cycles of chidamide plus R-GDP, 57.1% required a chidamide dose reduction. The most frequent treatment-emergent adverse event was anemia, occurring in all patients. Thrombocytopenia was the most common grade ≥3 adverse event leading to treatment interruption or dose reduction, affecting 48.1% of patients. Non-hematologic events included hypocalcemia (59.3%), hyponatremia (55.6%), and hypokalemia (44.4%). Three patients experienced grade 3 pneumonitis, and one patient had grade 3 skin rash. Conclusion: Chidamide plus R-GDP salvage chemotherapy demonstrated promising efficacy (ORR: 79.2%; CR rate: 45.8%) in r/r DLBCL patients. Adverse events were manageable but necessitated monitoring. Larger trials are needed to validate these findings and establish chidamide plus R-GDP as a potential treatment option for this challenging population. Keywords: relapsed/refractory DLBCL, salvage chemotherapy, chidamide, R-GDP, efficacy, safety

Assessment of exposure-response relationship for cetuximab in patients with metastatic colorectal cancer and head and neck cancer

ObjectiveThere is limited scientific evidence on the cetuximab exposure-response relationship and no concentration threshold has been associated with optimal disease control. The aims were to assess, in a real-life setting, the relationship between steady state cetuximab concentrations (Ctrough, ss) and disease control. MethodA prospective observational study in patients with metastatic colorectal cancer or head and neck cancer treated with cetuximab. Steady state trough concentrations were compared with the results of radiological assessment of response (progression or clinical benefit). Generalized estimating equations analysis was performed. To test the association between steady state concentrations and overall survival and progression-free survival, Cox proportional hazard models were developed. An optimal cut-off point was searched using the area under the receiver operating characteristic curve. ResultsA total of 30 steady state cetuximab concentrations from 16 patients were analysed. Median Ctrough, ss was 26.86 mg/L and there was marked inter- and intraindividual variability (standard deviation 32.4 mg/L and 16.9 mg/L, respectively). A positive association was found between cetuximab Ctrough, ss and clinical benefit (odds ratio 1.24, 95% confidence interval: 0.95-1.63, p = 0.113), although without reaching statistical significance. The area under the receiver operating characteristic curve (n = 30) had moderate discrimination power (0.71; 95% confidence interval 0.49-0.93), and the empirical optimal cutoff point was 19.12 mg/L. However, no association was observed between cetuximab Ctrough, SS and survival in metastatic colorectal cancer or neck cancer patients. ConclusionsWe cannot confirm a relationship between cetuximab Ctrough, SS and disease control despite a positive association. This study was conducted with a small sample, which reduces the power analysis. Further controlled randomised studies with a sufficient number of patients are needed.

Randomized Controlled Trial of Neoadjuvant Short-Course Radiotherapy Followed by Consolidation Chemotherapy Versus Long-Course Chemoradiotherapy in Locally Advanced Rectal Cancer: Comparison of Overall Response Rates.

Management of locally advanced rectal cancer (LARC) is evolving with current emphasis on the addition of chemotherapy to short course radiotherapy (SCRT). We primarily aimed to analyse the difference in overall response rates between SCRT with sequential chemotherapy and standard long-course chemoradiotherapy (LCCRT)in LARC. After randomization, patients in arm A received 45Gy in 25 fractions over 5weeks with concurrent capecitabine while patients in arm B received 25Gy in 5 fractions over 1week followed by 3 cycles of CAPOX (capecitabine and oxaliplatin) chemotherapy. Clinical and radiological response assessment was made after the completion of neoadjuvant treatment, a week prior to surgery. Adjuvant chemotherapy was added to complete 6months of peri-operative chemotherapy. Surgery was performed between 8 and 10weeks of completion of radiation treatment in both arms. Of the 33 patients recruited in this study between February 2020 to July 2021, 17 patients were randomized to arm A and 16 to arm B. The rates of complete tumour regression were 23.1% in arm A versus 35.7% in arm B (p-value = 0.683). Pathological complete response (pCR) rate was 20% arm A versus 30% in arm B (0.446). A higher number of patients in arm B experienced grade 3 diarrhoea, whereas acute skin toxicity was seen only in arm A. SCRT had fewer treatment interruptions compared to LCCRT. SCRT followed by three cycles of CAPOX chemotherapy in the neoadjuvant setting is comparable to LCCRT in terms of tumour response. This may be a better alternative regimen with fewer treatment interruptions in a resource-limited setting.

Diagnostic accuracy of digital mammography and breast ultrasonography in prediction of response to neoadjuvant systemic treatment in breast cancer patients

Aim: The aim of our study was to assess the accuracy of a combination of digital mammography and breast ultrasonography in the prediction of response to neoadjuvant systemic treatment in breast cancer patients with different tumor subtypes. Methods: The study was designed as a retrospective diagnostic accuracy study. Stage I-III female breast cancer patients who received any type of neoadjuvant systemic treatment with radiological response assessment by both mammography and breast ultrasound and followed by surgical treatment in the breast and axilla were included in the study. The primary outcome was the diagnostic accuracy of combined modalities of mammography and breast ultrasonography for predicting the pathological complete response. On mammography and breast ultrasonography, the radiological response was categorized into complete response and non-complete response. Pathological complete response on surgical specimens was described based on current guidelines. True and false positive cases as well as true and false negative cases were counted and compared among patients with 4 different molecular subtypes. The diagnostic accuracy of combined imaging modalities was analyzed for positive and negative predictive values, sensitivity, and specificity rates. All rates were calculated according to the previously described formulas. Results: Eighty-one breast cancer cases were included in the study. Positive predictive values of imaging were 100%, 75%, 100%, and 83%, whereas negative predictive values were 67%, 75%, 100%, and 100% in patients with HR+/HER2-, HR+/HER2+, HR-/HER2+ and HR-/HER2- tumors, respectively. Sensitivity rates were found to be 98%, 90%, 100%, and 100%, whereas specificity rates were 100%, 50%, 100%, and 67% in patients with HR+/HER2-, HR+/HER2+, HR-/HER2+ and HR-/HER2- tumors, respectively. Conclusion: Digital mammography and breast ultrasonography as a combined modality seem to show the pathological complete response after neoadjuvant systemic treatment in HR+/HER2- breast cancer patients with a very high specificity rate. Therefore, these conventional tools may help surgeons to select patients who might benefit from loco-regional treatment de-escalation with higher accuracy.

A comparative study evaluating the quality of life in patients receiving chemotherapy versus oral TKI in the third line and beyond setting for advanced non–small-cell lung cancer.

e21195 Background: The outcomes in advanced non-small cell lung cancer (NSCLC) have improved due to the availability of more effective systemic &amp; improved supportive care. This has increased the number of patients who seek treatment in the third line &amp; beyond setting. We conducted this study to compare the quality of life (QoL), toxicity &amp; outcomes in patients receiving chemotherapy &amp; oral tyrosine kinase inhibitors (TKIs) in this setting. Methods: In this phase 3 randomized open-label study patients with stage III/IV NSCLC with disease progression on at least 2 prior lines of chemotherapy, with a life expectancy of at least 3 months, without prior TKI exposure, and stable brain metastases (if any) were included. Patients were randomised to receive chemotherapy(Gemcitabine/docetaxel/paclitaxel/vinorelbine) or TKI (erlotinib or gefitinib). Patients were evaluated at every visit for clinical benefit and toxicity (as per CTCAE v4.3). A radiological tumour response assessment was done every 8-12 weeks from the start of therapy. The QoL was assessed using the EORTC QLQ C-30 &amp; LC-13 questionnaires at baseline and every 8-10 weeks while on treatment. The primary endpoint is the change in QoL scores at 8-10 weeks, the secondary endpoints are safety and overall survival. The change in QoL scores at baseline and scores at 8-10 weeks was determined and the mean difference between the arms was compared using the independent t-test. Overall and progression-free survival was determined using the Kaplan-Meier method and Cox proportional regression analysis. Results 246 patients were enrolled in the study, 123 in each arm. There was a male predominance in both arms. There was no significant difference in the change in the QoL scores from baseline to the subsequent visit (at 8-10 weeks) in both arms in all domains of QLQ C-30 &amp; LC 13 except alopecia. The mean difference in scores for alopecia was 23.73 (+/-52.22) in the chemotherapy arm &amp; -18.35 (+/- 47.32) in the oral TKI arm (p = 0.000). The median follow-up was 88.1 mos (95%CI39.04-137.15). On ITT analysis the median progression-free survival (PFS) was 3.13 mos (95%CI 2.15-4.11) in the chemotherapy arm and 2.26 mos (95%CI 2.1-2.43) in the oral TKI arm, hazard ratio (HR) 1.074 (95% CI 0.833- 1.38), p = 0.6. There were 120 deaths in each arm. The median overall survival (OS) was 7.63 mos (95% CI 5.96-9.30) in the chemotherapy arm and 7.5 mos in the oral TKI arm (95% CI 5.85-9.14); HR 1.024 (95%CI 0.793-1.321),p = 0.9. The toxicity profile was similar in both arms except for CIPN, alopecia, pedal edema which was higher in the chemotherapy arm, &amp; dry skin &amp; skin rash was higher in the oral TKI arm Conclusion In the third line setting there is no significant difference in most QoL domains with similar outcomes( PFS&amp; OS) and toxicity profile is consistent with expected toxicities of the drugs.

Toward automating RECIST 1.1: Improving AI new lesion detection with longitudinal image data.

e13545 Background: Automating RECIST 1.1 would offer benefits in speed, consistency, and cost of clinical trial imaging reads. Detection of new lesions is subjective and a common source of reader adjudication. Current AI lesion detection typically uses single timepoint images. The use of temporal data in AI is a rapidly evolving field in non-medical applications and new lesion detection in response assessment may benefit from longitudinal multi timepoint images. We explore the impact of longitudinal image input to deep neural networks for new lesion detection; as a proof of concept, we compare single vs dual timepoint input in new liver lesion detection in CT. Methods: We utilized CT images from two public datasets: 1) diseased liver (DL), a subset of DeepLesion (doi: 10.1117/1.JMI.5.3.036501), and 2) healthy liver (HL) from potential liver donors (Med Image Anal 69 (2021) 101950).. Due to the lack of a publicly available longitudinal image dataset with new lesions, we simulated (“sim”) a longitudinal study by registering HL scans (sim healthy baseline) with DL scans (sim follow up with new lesion) and selecting well-paired images using mutual information. From each pair of scans, we select one positive (with new lesion on follow up) and one negative (without lesion on follow up) 6x6cm image patch from within the liver, creating a dataset of 3296 patches with a balanced number of pos/neg samples. We trained two deep neural networks to classify the presence of a new liver lesion: 1) a conventional single timepoint model (ST) trained on DL patches only, i.e. using the follow up scan only, and 2) an experimental dual-timepoint model (DT) trained on patch pairs from DL-HL paired scans. We used 5-fold cross validation and paired t tests to compare model AUC and accuracy. Results: Of 10,594 CT studies (4,427 patients, pts) in DeepLesion, we paired 1648 DL scans (946 pts with liver lesions) with one of 20 registered HL scans (20 pts). The DT model outperformed the ST model, increasing AUC from 0.914 ± 0.002 to 0.929 ± 0.005 (p=0.040), with similar trends in accuracy (dataset split, Table 1). Conclusions: The DT model outperformed the ST model in the detection of new liver lesions, suggesting preliminary the benefit of multi-timepoint input to a deep neural network using sim longitudinal data. The use of longitudinal image data to improve interval change detection is a step towards automating radiologic response assessment in clinical trials. We plan to extend this work to automate new lesion detection in real clinical trial longitudinal images. Results (bold) are reported on test set across 5 folds. N image patches (ST) or patch pairs (DT). [Table: see text]