An International Expert-Based Consensus on the Definition of a Clinical Near-Complete Response After Neoadjuvant (Chemo)radiotherapy for Rectal Cancer

BACKGROUND

A variety of definitions for a clinical near-complete response after neoadjuvant (chemo) radiotherapy for rectal cancer are currently used. This variety leads to inconsistency in clinical practice, long-term outcome, and trial enrollment.

OBJECTIVE

The aim of this study was to reach expert-based consensus on the definition of a clinical near-complete response after (chemo) radiotherapy.

DESIGN

A modified Delphi process, including a systematic review, 3 surveys, and 2 meetings, was performed with an international expert panel consisting of 7 surgeons and 4 radiologists. The surveys consisted of individual features, statements, and feature combinations (endoscopy, T2-weighted MRI, and diffusion-weighted MRI).

SETTING

The modified Delphi process was performed in an online setting; all 3 surveys were completed online by the expert panel, and both meetings were hosted online.

MAIN OUTCOME MEASURES

The main outcome was to reach consensus (80% or more agreement).

RESULTS

The expert panel reached consensus on a 3-tier categorization of the near-complete response category based on the likelihood of the response to evolve into a clinical complete response after a longer waiting interval. The panelists agreed that a near-complete response is a temporary entity only to be used in the first 6 months after (chemo)radiotherapy. Furthermore, consensus was reached that the lymph node status should be considered when deciding on a near-complete response and that biopsies are not always needed when a near-complete response is found. No consensus was reached on whether primary staging characteristics have to be taken into account when deciding on a near-complete response.

LIMITATIONS

This 3-tier subcategorization is expert-based; therefore, there is no supporting evidence for this subcategorization. Also, it is unclear whether this subcategorization can be generalized into clinical practice.

CONCLUSIONS

Consensus was reached on the use of a 3-tier categorization of a near-complete response, which can be helpful in daily practice as guidance for treatment and to inform patients with a near-complete response on the likelihood of successful organ preservation. See Video Abstract.

UN CONSENSO INTERNACIONAL BASADO EN EXPERTOS ACERCA DE LA DEFINICIN DE UNA RESPUESTA CLNICA CASI COMPLETA DESPUS DE QUIMIORADIOTERAPIA NEOADYUVANTE CONTRA EL CNCER DE RECTO

ANTECEDENTES:Actualmente, se utilizan una variedad de definiciones para una respuesta clínica casi completa después de quimioradioterapia neoadyuvante contra el cáncer de recto. Esta variedad resulta en inconsistencia en la práctica clínica, los resultados a largo plazo y la inscripción en ensayos.OBJETIVO:El objetivo de este estudio fue llegar a un consenso de expertos sobre la definición de una respuesta clínica casi completa después de quimioradioterapia.DISEÑO:Se realizó un proceso Delphi modificado que incluyó una revisión sistemática, 3 encuestas y 2 reuniones con un panel internacional de expertos compuesto por siete cirujanos y 4 radiólogos. Las encuestas consistieron en características individuales, declaraciones y combinaciones de características (endoscopía, T2W-MRI y DWI).AJUSTE:El proceso Delphi modificado se realizó en un entorno en línea; el panel de expertos completó las tres encuestas en línea y ambas reuniones se realizaron en línea.PRINCIPALES MEDIDAS DE RESULTADO:El resultado principal fue llegar a un consenso (≥80% de acuerdo).RESULTADOS:El panel de expertos llegó a un consenso sobre una categorización de tres niveles de la categoría de respuesta casi completa basada en la probabilidad de que la respuesta evolucione hacia una respuesta clínica completa después de un intervalo de espera más largo. Los panelistas coincidieron en que una respuesta casi completa es una entidad temporal que sólo debe utilizarse en los primeros 6 meses después de la quimioradioterapia. Además, se llegó a un consenso en que se debe considerar el estado de los nódulos linfáticos al decidir sobre una respuesta casi completa y que no siempre se necesitan biopsias cuando se encuentra una respuesta casi completa. No se llegó a un consenso sobre si se deben tener en cuenta las características primarias de estadificación al decidir una respuesta casi completa.LIMITACIONES:Esta subcategorización de 3 niveles está basada en expertos; por lo tanto, no hay evidencia que respalde esta subcategorización. Además, no está claro si esta subcategorización puede generalizarse a la práctica clínica.CONCLUSIONES:Se alcanzó consenso sobre el uso de una categorización de 3 niveles de una respuesta casi completa que puede ser útil en la práctica diaria como guía para el tratamiento y para informar a los pacientes con una respuesta casi completa sobre la probabilidad de una preservación exitosa del órgano. (Traducción - Dr. Aurian Garcia Gonzalez).

How Does Target Lesion Selection Affect RECIST? A Computer Simulation Study

OBJECTIVES

Response Evaluation Criteria in Solid Tumors (RECIST) is grounded on the assumption that target lesion selection is objective and representative of the change in total tumor burden (TTB) during therapy. A computer simulation model was designed to challenge this assumption, focusing on a particular aspect of subjectivity: target lesion selection.

MATERIALS AND METHODS

Disagreement among readers and the disagreement between individual reader measurements and TTB were analyzed as a function of the total number of lesions, affected organs, and lesion growth.

RESULTS

Disagreement rises when the number of lesions increases, when lesions are concentrated on a few organs, and when lesion growth borders the thresholds of progressive disease and partial response. There is an intrinsic methodological error in the estimation of TTB via RECIST 1.1, which depends on the number of lesions and their distributions. For example, for a fixed number of lesions at 5 and 15, distributed over a maximum of 4 organs, the error rates are observed to be 7.8% and 17.3%, respectively.

CONCLUSIONS

Our results demonstrate that RECIST can deliver an accurate estimate of TTB in localized disease, but fails in cases of distal metastases and multiple organ involvement. This is worsened by the "selection of the largest lesions," which introduces a bias that makes it hardly possible to perform an accurate estimate of the TTB. Including more (if not all) lesions in the quantitative analysis of tumor burden is desirable.

Prognostic significance of MRI-detected extramural venous invasion according to grade and response to neo-adjuvant treatment in locally advanced rectal cancer A national cohort study after radiologic training and reassessment

BACKGROUND

Detection of grade 3-4 extra mural venous invasion (mrEMVI) on magnetic resonance imaging (MRI) is associated with an increased distant metastases (DM)-rate. This study aimed to determine the impact of different grades of mrEMVI and their disappearance after neoadjuvant therapy.

METHODS

A Dutch national retrospective cross-sectional study was conducted, including patients who underwent resection for rectal cancer in 2016 from 60/69 hospitals performing rectal surgery. Patients with a cT3-4 tumour ≤8 cm from the anorectal junction were selected and their MRI-scans were reassessed by trained abdominal radiologists. Positive mrEMVI grades (3 and 4) were analyzed in regard to 4-year local recurrence (LR), DM, disease-free survival (DFS) and overall survival (OS).

RESULTS

The 1213 included patients had a median follow-up of 48 months (IQR 30-54). Positive mrEMVI was present in 324 patients (27%); 161 had grade 3 and 163 had grade 4. A higher mrEMVI stage (grade 4 vs grade 3 vs no mrEMVI) increased LR-risk (21% vs 18% vs 7%, <0.001) and DM-risk (49% vs 30% vs 21%, p < 0.001) and decreased DFS (42% vs 55% vs 69%, p < 0.001) and OS (62% vs 76% vs 81%, p < 0.001), which remained independently associated in multivariable analysis. When mrEMVI had disappeared on restaging MRI, DM-rate was comparable to initial absence of mrEMVI (both 26%), whereas LR-rate remained high (22% vs 9%, p = 0.006).

CONCLUSION

The negative oncological impact of mrEMVI on recurrence and survival rates was dependent on grading. Disappearance of mrEMVI on restaging MRI decreased the risk of DM, but not of LR.

Evolution of clinical nature, treatment and survival of locally recurrent rectal cancer: Comparative analysis of two national cross-sectional cohorts

BACKGROUND

In the Netherlands, use of neoadjuvant radiotherapy for rectal cancer declined after guideline revision in 2014. This decline is thought to affect the clinical nature and treatability of locally recurrent rectal cancer (LRRC). Therefore, this study compared two national cross-sectional cohorts before and after the guideline revision with the aim to determine the changes in treatment and survival of LRRC patients over time.

METHODS

Patients who underwent resection of primary rectal cancer in 2011 (n = 2094) and 2016 (n = 2855) from two nationwide cohorts with a 4-year follow up were included. Main outcomes included time to LRRC, synchronous metastases at time of LRRC diagnosis, intention of treatment and 2-year overall survival after LRRC.

RESULTS

Use of neoadjuvant (chemo)radiotherapy for the primary tumour decreased from 88.5% to 60.0% from 2011 to 2016. The 3-year LRRC rate was not significantly different with 5.1% in 2011 (n = 114, median time to LRRC 16 months) and 6.3% in 2016 (n = 202, median time to LRRC 16 months). Synchronous metastasis rate did not significantly differ (27.2% vs 33.7%, p = 0.257). Treatment intent of the LRRC shifted towards more curative treatment (30.4% vs. 47.0%, p = 0.009). In the curatively treated group, two-year overall survival after LRRC diagnoses increased from 47.5% to 78.7% (p = 0.013).

CONCLUSION

Primary rectal cancer patients in 2016 were treated less often with neoadjuvant (chemo)radiotherapy, while LRRC rates remained similar. Those who developed LRRC were more often candidate for curative intent treatment compared to the 2011 cohort, and survival after curative intent treatment also improved substantially.

Overcoming data scarcity in radiomics/radiogenomics using synthetic radiomic features

PURPOSE

To evaluate the potential of synthetic radiomic data generation in addressing data scarcity in radiomics/radiogenomics models.

METHODS

This study was conducted on a retrospectively collected cohort of 386 colorectal cancer patients (n = 2570 lesions) for whom matched contrast-enhanced CT images and gene TP53 mutational status were available. The full cohort data was divided into a training cohort (n = 2055 lesions) and an independent and fixed test set (n = 515 lesions). Differently sized training sets were subsampled from the training cohort to measure the impact of sample size on model performance and assess the added value of synthetic radiomic augmentation at different sizes. Five different tabular synthetic data generation models were used to generate synthetic radiomic data based on "real-world" radiomics data extracted from this cohort. The quality and reproducibility of the generated synthetic radiomic data were assessed. Synthetic radiomics were then combined with "real-world" radiomic training data to evaluate their impact on the predictive model's performance.

RESULTS

A prediction model was generated using only "real-world" radiomic data, revealing the impact of data scarcity in this particular data set through a lack of predictive performance at low training sample numbers (n = 200, 400, 1000 lesions with average AUC = 0.52, 0.53, and 0.56 respectively, compared to 0.64 when using 2055 training lesions). Synthetic tabular data generation models created reproducible synthetic radiomic data with properties highly similar to "real-world" data (for n = 1000 lesions, average Chi-square = 0.932, average basic statistical correlation = 0.844). The integration of synthetic radiomic data consistently enhanced the performance of predictive models trained with small sample size sets (AUC enhanced by 9.6%, 11.3%, and 16.7% for models trained on n_samples = 200, 400, and 1000 lesions, respectively). In contrast, synthetic data generated from randomised/noisy radiomic data failed to enhance predictive performance underlining the requirement of true signal data to do so.

CONCLUSION

Synthetic radiomic data, when combined with real radiomics, could enhance the performance of predictive models. Tabular synthetic data generation might help to overcome limitations in medical AI stemming from data scarcity.

Multi-omics staging of locally advanced rectal cancer predicts treatment response: a pilot study

Treatment response assessment of rectal cancer patients is a critical component of personalized cancer care and it allows to identify suitable candidates for organ-preserving strategies. This pilot study employed a novel multi-omics approach combining MRI-based radiomic features and untargeted metabolomics to infer treatment response at staging. The metabolic signature highlighted how tumor cell viability is predictively down-regulated, while the response to oxidative stress was up-regulated in responder patients, showing significantly reduced oxoproline values at baseline compared to non-responder patients (p-value < 10-4). Tumors with a high degree of texture homogeneity, as assessed by radiomics, were more likely to achieve a major pathological response (p-value < 10-3). A machine learning classifier was implemented to summarize the multi-omics information and discriminate responders and non-responders. Combining all available radiomic and metabolomic features, the classifier delivered an AUC of 0.864 (± 0.083, p-value < 10-3) with a best-point sensitivity of 90.9% and a specificity of 81.8%. Our results suggest that a multi-omics approach, integrating radiomics and metabolomic data, can enhance the predictive value of standard MRI and could help to avoid unnecessary surgical treatments and their associated long-term complications.

Impact of the new rectal cancer definition on multimodality treatment and interhospital variability: Results from a nationwide cross-sectional study

AIM

This study aimed to determine the consequences of the new definition of rectal cancer for decision-making in multidisciplinary team meetings (MDT). The new definition of rectal cancer, the lower border of the tumour is located below the sigmoid take-off (STO), was implemented in the Dutch guideline in 2019 after an international Delphi consensus meeting to reduce interhospital variations.

METHOD

All patients with rectal cancer according to the local MDT, who underwent resection in 2016 in the Netherlands were eligible for this nationwide collaborative cross-sectional study. MRI-images were rereviewed, and the tumours were classified as above or on/below the STO.

RESULTS

This study registered 3107 of the eligible 3178 patients (98%), of which 2784 patients had an evaluable MRI. In 314 patients, the tumour was located above the STO (11%), with interhospital variation between 0% and 36%. Based on TN-stage, 175 reclassified patients with colon cancer (6%) would have received different treatment (e.g., omitting neoadjuvant radiotherapy, candidate for adjuvant chemotherapy). Tumour location above the STO was independently associated with lower risk of 4-year locoregional recurrence (HR 0.529; p = 0.030) and higher 4-year overall survival (HR 0.732; p = 0.037) compared to location under the STO.

CONCLUSION

By using the STO, 11% of the prior MDT-based diagnosis of rectal cancer were redefined as sigmoid cancer, with potential implications for multimodality treatment and prognostic value. Given the substantial interhospital variation in proportion of redefined cancers, the use of the STO will contribute to standardisation and comparability of outcomes in both daily practice and trial settings.

Association of Lateral Pelvic Lymph Nodes with Disease Recurrence and Organ Preservation in Patients with Distal Rectal Adenocarcinoma Treated with Total Neoadjuvant Therapy

OBJECTIVE

Assess the significance of enlarged lateral lymph nodes (LLN) for disease recurrence, metastasis, and organ preservation in patients with rectal cancer.

BACKGROUND

Optimal treatment of rectal adenocarcinoma involving LLN is subject to debate.

METHODS

A post hoc analysis of the OPRA trial, a multicenter study of patients with rectal cancer treated with total neoadjuvant therapy (TNT) followed by total mesorectal excision or watch-and-wait management. We analyzed the association of visible LLN (LLN+), LLN≥7 mm (short axis) on baseline MRI, and LLN≥4 mm on restaging MRI with recurrence, metastasis, and rectum preservation.

RESULTS

At baseline, 57 out of 324 (18%) patients had LLN+. In 30 (53%) of 57 patients with LLN+ on baseline MRI, the LLN disappeared after TNT. Disease recurrence in LLN was rare (3.5% of patients with LLN+ and 0.4% of patients with LLN-). All patients with recurrence in LLN also had distant metastasis. The rate of organ preservation was significantly lower in patients with LLN≥4 mm on restaging MRI (P=0.013). We found no significant differences in rates of local recurrence or metastasis between patients with LLN+ vs. LLN- and in patients with LLN≥7 vs.<7 mm on baseline MRI. LLN dissection was performed in 3 patients; 2 of them died of distant metastasis.

CONCLUSIONS

LLN involvement is not associated with disease recurrence or metastasis, but persistence of LLN≥4 mm after TNT is negatively associated with rectum preservation in patients with locally advanced rectal cancer treated with TNT. Dissection of lateral nodes likely benefits few patients.

Imaging in the era of risk-adapted treatment in Colon cancer

The treatment landscape for patients with colon cancer is continuously evolving. Risk-adapted treatment strategies, including neoadjuvant chemotherapy and immunotherapy, are slowly finding their way into clinical practice and guidelines. Radiologists are pivotal in guiding clinicians toward the most optimal treatment for each colon cancer patient. This review provides an overview of recent and upcoming advances in the diagnostic management of colon cancer and the radiologist's role in the multidisciplinary approach to treating colon cancer.

Investigating locations of recurrences with MRI after CRS-HIPEC for colorectal peritoneal metastases

PURPOSE

Patients with colorectal peritoneal metastases (PM) treated with cytoreductive surgery (CRS) and hyperthermic intraperitoneal chemotherapy (HIPEC) are at high risk of recurrent disease. Understanding where and why recurrences occur is the first step in finding solutions to reduce recurrence rates. Although diffusion-weighted (DW) MRI is not routinely used in the follow-up of CRC patients, it has a clear advantage over CT in detecting the location and spread of (recurrent) PM. This study aimed to identify common locations of recurrence in CRC patients after CRS-HIPEC with MRI.

METHOD

This was a single-centre retrospective study of patients with recurrent PM after CRS-HIPEC performed between January 2016 and August 2020. Patients were eligible for inclusion if they had both an MRI preoperatively (MRI1) and at the time of recurrent disease (MRI2). Two abdominal radiologists reviewed in consensus and categorized recurrences according to their location on MRI2 and in correlation with previous disease location on prior imaging (MRI1) and the surgical report of the CRS-HIPEC.

RESULTS

Thirty patients were included, with a median surgical PCI of 7 (range 3-21) at the time of primary CRS-HIPEC. In total, 68 recurrent metastases were detected on MRI2, of which 14 were extra-peritoneal. Of the remaining 54 PM, 42 (78%) occurred where the peritoneum was damaged due to earlier resections or other surgical procedures (e.g. inserted surgical abdominal drains). Most recurrent metastases were found in the mesentery, lower abdomen/pelvis and abdominal wall (87%).

CONCLUSIONS

Most recurrent PMs appeared in the mesentery, lower abdomen/pelvis and abdominal wall, especially where the peritoneum was previously damaged.

Artificial intelligence-assisted double reading of chest radiographs to detect clinically relevant missed findings: a two-centre evaluation

OBJECTIVES

To evaluate an artificial intelligence (AI)-assisted double reading system for detecting clinically relevant missed findings on routinely reported chest radiographs.

METHODS

A retrospective study was performed in two institutions, a secondary care hospital and tertiary referral oncology centre. Commercially available AI software performed a comparative analysis of chest radiographs and radiologists' authorised reports using a deep learning and natural language processing algorithm, respectively. The AI-detected discrepant findings between images and reports were assessed for clinical relevance by an external radiologist, as part of the commercial service provided by the AI vendor. The selected missed findings were subsequently returned to the institution's radiologist for final review.

RESULTS

In total, 25,104 chest radiographs of 21,039 patients (mean age 61.1 years ± 16.2 [SD]; 10,436 men) were included. The AI software detected discrepancies between imaging and reports in 21.1% (5289 of 25,104). After review by the external radiologist, 0.9% (47 of 5289) of cases were deemed to contain clinically relevant missed findings. The institution's radiologists confirmed 35 of 47 missed findings (74.5%) as clinically relevant (0.1% of all cases). Missed findings consisted of lung nodules (71.4%, 25 of 35), pneumothoraces (17.1%, 6 of 35) and consolidations (11.4%, 4 of 35).

CONCLUSION

The AI-assisted double reading system was able to identify missed findings on chest radiographs after report authorisation. The approach required an external radiologist to review the AI-detected discrepancies. The number of clinically relevant missed findings by radiologists was very low.

CLINICAL RELEVANCE STATEMENT

The AI-assisted double reader workflow was shown to detect diagnostic errors and could be applied as a quality assurance tool. Although clinically relevant missed findings were rare, there is potential impact given the common use of chest radiography.

KEY POINTS

• A commercially available double reading system supported by artificial intelligence was evaluated to detect reporting errors in chest radiographs (n=25,104) from two institutions. • Clinically relevant missed findings were found in 0.1% of chest radiographs and consisted of unreported lung nodules, pneumothoraces and consolidations. • Applying AI software as a secondary reader after report authorisation can assist in reducing diagnostic errors without interrupting the radiologist's reading workflow. However, the number of AI-detected discrepancies was considerable and required review by a radiologist to assess their relevance.

Multi-sequence MRI radiomics of colorectal liver metastases: Which features are reproducible across readers?

PURPOSE

To assess the inter-reader reproducibility of radiomics features on multiple MRI sequences after segmentations of colorectal liver metastases (CRLM).

METHOD

30 CRLM (in 23 patients) were manually delineated by three readers on MRI before the start of chemotherapy on the contrast enhanced T1-weighted images (CE-T1W) in the portal venous phase, T2-weighted images (T2W) and b800 diffusion weighted images (DWI). DWI delineations were copied to the ADC-maps. 107 radiomics features were extracted per sequence. The intraclass correlation coefficient (ICC) was calculated per feature. Features were considered reproducible if ICC > 0.9.

RESULTS

90% of CE-T1W features were reproducible with a median ICC of 0.98 (range 0.76-1.00). 81% of DWI features were robust with median ICC = 0.97 (range 0.38-1.00). The T2W features had a median ICC of 0.96 (range 0.55-0.99) and were reproducible in 80%. ADC showed the lowest number of reproducible features with 58% and median ICC = 0.91 (range 0.38-0.99) When considering the lower bound of the ICC 95% confidence intervals, 58%, 66%, 54% and 29% reached 0.9 for the CE-T1W, DWI, T2W and ADC features, respectively. The feature class with the best reproducibility differed per sequence.

CONCLUSIONS

The majority of MRI radiomics features from CE-T1W, T2W, DWI and ADC in colorectal liver metastases were robust for segmentation variability between readers. The CE-T1W yielded slightly better reproducibility results compared to DWI and T2W. The ADC features seem more susceptible to reader differences compared to the other three sequences.

Outcomes and potential impact of a virtual hands-on training program on MRI staging confidence and performance in rectal cancer

OBJECTIVES

To explore the potential impact of a dedicated virtual training course on MRI staging confidence and performance in rectal cancer.

METHODS

Forty-two radiologists completed a stepwise virtual training course on rectal cancer MRI staging composed of a pre-course (baseline) test with 7 test cases (5 staging, 2 restaging), a 1-day online workshop, 1 month of individual case readings (n = 70 cases with online feedback), a live online feedback session supervised by two expert faculty members, and a post-course test. The ESGAR structured reporting templates for (re)staging were used throughout the course. Results of the pre-course and post-course test were compared in terms of group interobserver agreement (Krippendorf's alpha), staging confidence (perceived staging difficulty), and diagnostic accuracy (using an expert reference standard).

RESULTS

Though results were largely not statistically significant, the majority of staging variables showed a mild increase in diagnostic accuracy after the course, ranging between + 2% and + 17%. A similar trend was observed for IOA which improved for nearly all variables when comparing the pre- and post-course. There was a significant decrease in the perceived difficulty level (p = 0.03), indicating an improved diagnostic confidence after completion of the course.

CONCLUSIONS

Though exploratory in nature, our study results suggest that use of a dedicated virtual training course and web platform has potential to enhance staging performance, confidence, and interobserver agreement to assess rectal cancer on MRI virtual training and could thus be a good alternative (or addition) to in-person training.

CLINICAL RELEVANCE STATEMENT

Rectal cancer MRI reporting quality is highly dependent on radiologists' expertise, stressing the need for dedicated training/teaching. This study shows promising results for a virtual web-based training program, which could be a good alternative (or addition) to in-person training.

KEY POINTS

• Rectal cancer MRI reporting quality is highly dependent on radiologists' expertise, stressing the need for dedicated training and teaching. • Using a dedicated virtual training course and web-based platform, encouraging first results were achieved to improve staging accuracy, diagnostic confidence, and interobserver agreement. • These exploratory results suggest that virtual training could thus be a good alternative (or addition) to in-person training.

Whole-body MRI with diffusion-weighted imaging as an adjunct to18 F-fluorodeoxyglucose positron emission tomography and CT in patients with suspected recurrent colorectal cancer

AIM

The aim was to explore how findings of whole-body MRI including diffusion-weighted imaging (DW-MRI) compared to the routine diagnostic workup with CT and/or 18 F-fluorodeoxyglucose (FDG) positron emission tomography (PET)/CT in patients with suspected recurrent colorectal cancer (CRC).

METHOD

This was an exploratory retrospective analysis of 55 patients with a clinical suspicion of recurrent CRC who underwent DW-MRI following CT and/or FDG-PET/CT. Two readers in consensus interpreted all clinical imaging reports and converted each described lesion into a confidence score (1 = definitely benign to 5 = definitely malignant). DW-MRI findings were compared to the most recent previous CT or PET/CT. Any discrepant or additional DW-MRI findings were documented and compared with histology and/or clinical follow-up (if available).

RESULTS

Whole-body MRI including diffusion-weighted imaging (DW-MRI) resulted in discrepant/additional findings in 26/55 (47%) cases; 23/37 (62%) compared to previous CT and 3/18 (17%) compared to previous PET/CT. These included 10 cases where DW-MRI converted previously inconclusive CT (n = 8) or PET/CT (n = 2) findings into a conclusive diagnosis, one where it contradicted a previous CT diagnosis of recurrence, five where DW-MRI diagnosed recurrent disease not previously reported on CT and 10 cases where DW-MRI detected additional lesions compared to CT (n = 9) or PET/CT (n = 1). Eighty-eight per cent of cases with discrepant/additional findings concerned patients with recurrent/metachronous peritoneal metastases. In total, DW-MRI resulted in 42 discrepant/additional lesions; the DW-MRI diagnosis was correct in 76% of these lesions and incorrect (false positive) in 7%. In the remaining 17%, no standard of reference was available.

CONCLUSIONS

This explorative study suggests that DW-MRI may be of added value to patients with a clinical suspicion for recurrent CRC, in particular to identify patients with peritoneal metastases. DW-MRI mainly has potential as a 'problem-solver' in patients with inconclusive or negative findings on previous imaging (in particular CT) and to detect additional disease sites in patients already diagnosed with recurrent disease.

The diagnostic accuracy of local staging in colon cancer based on computed tomography (CT): evaluating the role of extramural venous invasion and tumour deposits

PURPOSE

The shift from adjuvant to neoadjuvant treatment in colon cancer demands the radiological selection of patients for systemic therapy. The aim of this study was to evaluate the accuracy of the CT-based TNM stage and high-risk features, including extramural venous invasion (EMVI) and tumour deposits, in the identification of patients with histopathological advanced disease, currently considered for neoadjuvant treatment (T3-4 disease).

METHODS

All consecutive patients surgically treated for non-metastatic colon cancer between January 2018 and January 2020 in a referral centre for colorectal cancer were identified retrospectively. All tumours were staged on CT according to the TNM classification system. Additionally, the presence of EMVI and tumour deposits on CT was evaluated. The histopathological TNM classification was used as reference standard.

RESULTS

A total of 176 patients were included. Histopathological T3-4 colon cancer was present in 85.0% of the patients with CT-detected T3-4 disease. Histopathological T3-4 colon cancer was present in 96.4% of the patients with CT-detected T3-4 colon cancer in the presence of both CT-detected EMVI and CT-detected tumour deposits. Histopathological T0-2 colon cancer was present in 50.8% of the patients with CT-detected T0-2 disease, and in 32.4% of the patients without CT-detected EMVI and tumour deposits.

CONCLUSION

The diagnostic accuracy of CT-based staging was comparable with previous studies. The presence of high-risk features on CT increased the probability of histopathological T3-4 colon cancer. However, a substantial part of the patients without CT-detected EMVI and tumour deposits was diagnosed with histopathological T3-4 disease. Hence, more accurate selection criteria are required to correctly identify patients with locally advanced disease.

Abandonment of Routine Radiotherapy for Nonlocally Advanced Rectal Cancer and Oncological Outcomes

Importance

Neoadjuvant short-course radiotherapy was routinely applied for nonlocally advanced rectal cancer (cT1-3N0-1M0 with >1 mm distance to the mesorectal fascia) in the Netherlands following the Dutch total mesorectal excision trial. This policy has shifted toward selective application after guideline revision in 2014.

Objective

To determine the association of decreased use of neoadjuvant radiotherapy with cancer-related outcomes and overall survival at a national level.

Design, Setting, and Participants

This multicenter, population-based, nationwide cross-sectional cohort study analyzed Dutch patients with rectal cancer who were treated in 2011 with a 4-year follow-up. A similar study was performed in 2021, analyzing all patients that were surgically treated in 2016. From these cohorts, all patients with cT1-3N0-1M0 rectal cancer and radiologically unthreatened mesorectal fascia were included in the current study. The data of the 2011 cohort were collected between May and October 2015, and the data of the 2016 cohort were collected between October 2020 and November 2021. The data were analyzed between May and October 2022.

Main Outcomes and Measures

The main outcomes were 4-year local recurrence and overall survival rates.

Results

Among the 2011 and 2016 cohorts, 1199 (mean [SD] age, 68 [11] years; 430 women [36%]) of 2095 patients (57.2%) and 1576 (mean [SD] age, 68 [10] years; 547 women [35%]) of 3057 patients (51.6%) had cT1-3N0-1M0 rectal cancer and were included, with proportions of neoadjuvant radiotherapy of 87% (2011) and 37% (2016). Four-year local recurrence rates were 5.8% and 5.5%, respectively (P = .99). Compared with the 2011 cohort, 4-year overall survival was significantly higher in the 2016 cohort (79.6% vs 86.4%; P < .001), with lower non-cancer-related mortality (13.8% vs 6.3%; P < .001).

Conclusions and Relevance

The results of this cross-sectional study suggest that an absolute 50% reduction in radiotherapy use for nonlocally advanced rectal cancer did not compromise cancer-related outcomes at a national level. Optimizing clinical staging and surgery following the Dutch total mesorectal excision trial has potentially enabled safe deintensification of treatment.

Value of Size and Malignant Features of Lateral Lymph Nodes in Risk Stratification at Lateral Local Recurrence of Rectal Cancer: A National Cohort Study

BACKGROUND

Patients with rectal cancer who have enlarged lateral lymph nodes (LLNs) have an increased risk of lateral local recurrence (LLR). However, little is known about prognostic implications of malignant features (internal heterogeneity, irregular margins, loss of fatty hilum, and round shape) on MRI and number of enlarged LLNs, in addition to LLN size.

METHODS

Of the 3,057 patients with rectal cancer included in this national, retrospective, cross-sectional cohort study, 284 with a cT3-4 tumor located ≤8 cm from the anorectal junction who received neoadjuvant treatment and who had visible LLNs on MRI were selected. Imaging was reassessed by trained radiologists. LLNs were categorized based on size. Influence of malignant features and the number of LLNs on LLR was investigated.

RESULTS

Of 284 patients with at least 1 visible LLN, 122 (43%) had an enlarged node (≥7.0 mm) and 157 (55%) had malignant features. Of the 122 patients with enlarged nodes, 25 had multiple (≥2). In patients with a single enlarged node (n=97), a single malignant feature was associated with a 4-year LLR rate of 0% and multiple malignant features was associated with a rate of 17% (P=.060). In the group with multiple malignant features, their disappearance on restaging was associated with an LLR rate of 13% compared with an LLR rate of 20% for persistent malignant features (P=.532). The presence of intermediate-size LLNs (5.0-6.9 mm) with at least 1 malignant feature was associated with a 4-year LLR rate of 8%; the 4-year LLR rate was 13% when the malignant features persisted on restaging MRI (P=.409). Patients with multiple enlarged LLNs had a 4-year LLR rate of 28% compared with 11% for those with a single enlarged LLN (P=.059).

CONCLUSIONS

The presence of multiple enlarged LLNs (≥7.0 mm), as well as multiple malignant features in an enlarged node contribute to the risk of developing an LLR. These radiologic features can be used for clinical decision-making regarding the potential benefit of LLN dissection.

An automated deep learning pipeline for EMVI classification and response prediction of rectal cancer using baseline MRI: a multi-centre study

The classification of extramural vascular invasion status using baseline magnetic resonance imaging in rectal cancer has gained significant attention as it is an important prognostic marker. Also, the accurate prediction of patients achieving complete response with primary staging MRI assists clinicians in determining subsequent treatment plans. Most studies utilised radiomics-based methods, requiring manually annotated segmentation and handcrafted features, which tend to generalise poorly. We retrospectively collected 509 patients from 9 centres, and proposed a fully automated pipeline for EMVI status classification and CR prediction with diffusion weighted imaging and T2-weighted imaging. We applied nnUNet, a self-configuring deep learning model, for tumour segmentation and employed learned multiple-level image features to train classification models, named MLNet. This ensures a more comprehensive representation of the tumour features, in terms of both fine-grained detail and global context. On external validation, MLNet, yielding similar AUCs as internal validation, outperformed 3D ResNet10, a deep neural network with ten layers designed for analysing spatiotemporal data, in both CR and EMVI tasks. For CR prediction, MLNet showed better results than the current state-of-the-art model using imaging and clinical features in the same external cohort. Our study demonstrated that incorporating multi-level image representations learned by a deep learning based tumour segmentation model on primary MRI improves the results of EMVI classification and CR prediction with good generalisation to external data. We observed variations in the contributions of individual feature maps to different classification tasks. This pipeline has the potential to be applied in clinical settings, particularly for EMVI classification.

METhodological RadiomICs Score (METRICS): a quality scoring tool for radiomics research endorsed by EuSoMII

PURPOSE

To propose a new quality scoring tool, METhodological RadiomICs Score (METRICS), to assess and improve research quality of radiomics studies.

METHODS

We conducted an online modified Delphi study with a group of international experts. It was performed in three consecutive stages: Stage#1, item preparation; Stage#2, panel discussion among EuSoMII Auditing Group members to identify the items to be voted; and Stage#3, four rounds of the modified Delphi exercise by panelists to determine the items eligible for the METRICS and their weights. The consensus threshold was 75%. Based on the median ranks derived from expert panel opinion and their rank-sum based conversion to importance scores, the category and item weights were calculated.

RESULT

In total, 59 panelists from 19 countries participated in selection and ranking of the items and categories. Final METRICS tool included 30 items within 9 categories. According to their weights, the categories were in descending order of importance: study design, imaging data, image processing and feature extraction, metrics and comparison, testing, feature processing, preparation for modeling, segmentation, and open science. A web application and a repository were developed to streamline the calculation of the METRICS score and to collect feedback from the radiomics community.

CONCLUSION

In this work, we developed a scoring tool for assessing the methodological quality of the radiomics research, with a large international panel and a modified Delphi protocol. With its conditional format to cover methodological variations, it provides a well-constructed framework for the key methodological concepts to assess the quality of radiomic research papers.

CRITICAL RELEVANCE STATEMENT

A quality assessment tool, METhodological RadiomICs Score (METRICS), is made available by a large group of international domain experts, with transparent methodology, aiming at evaluating and improving research quality in radiomics and machine learning.

KEY POINTS

• A methodological scoring tool, METRICS, was developed for assessing the quality of radiomics research, with a large international expert panel and a modified Delphi protocol. • The proposed scoring tool presents expert opinion-based importance weights of categories and items with a transparent methodology for the first time. • METRICS accounts for varying use cases, from handcrafted radiomics to entirely deep learning-based pipelines. • A web application has been developed to help with the calculation of the METRICS score ( https://metricsscore.github.io/metrics/METRICS.html ) and a repository created to collect feedback from the radiomics community ( https://github.com/metricsscore/metrics ).

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

OBJECTIVES

To 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.

METHODS

The 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.

RESULTS

A 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%).

CONCLUSION

Since 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 STATEMENT

Compared 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.

An updated evaluation of the implementation of the sigmoid take-off landmark 1 year after the official introduction in the Netherlands

PURPOSE

The definition of rectal cancer based on the sigmoid take-off (STO) was incorporated into the Dutch guideline in 2019, and became mandatory in the national audit from December 2020. This study aimed to evaluate the use of the STO in clinical practice and the added value of online training, stratified for the period before (group A, historical cohort) and after (group B, current cohort) incorporation into the national audit.

METHODS

Participants, including radiologists, surgeons, surgical and radiological residents, interns, PhD students, and physician assistants, were asked to complete an online training program, consisting of questionnaires, 20 MRI cases, and a training document. Outcomes were agreement with the expert reference, inter-rater variability, and accuracy before and after the training.

RESULTS

Group A consisted of 86 participants and group B consisted of 114 participants. Familiarity with the STO was higher in group B (76% vs 88%, p = 0.027). Its use in multidisciplinary meetings was not significantly higher (50% vs 67%, p = 0.237). Agreement with the expert reference was similar for both groups before (79% vs 80%, p = 0.423) and after the training (87% vs 87%, p = 0.848). Training resulted in significant improvement for both groups in classifying tumors located around the STO (group A, 69-79%; group B, 67-79%, p < 0.001).

CONCLUSIONS

The results of this study show that after the inclusion of the STO in the mandatory Dutch national audit, the STO was consequently used in only 67% of the represented hospitals. Online training has the potential to improve implementation and unambiguous assessment.

Development and multicenter validation of a multiparametric imaging model to predict treatment response in rectal cancer

OBJECTIVES

To develop and validate a multiparametric model to predict neoadjuvant treatment response in rectal cancer at baseline using a heterogeneous multicenter MRI dataset.

METHODS

Baseline staging MRIs (T2W (T2-weighted)-MRI, diffusion-weighted imaging (DWI) / apparent diffusion coefficient (ADC)) of 509 patients (9 centres) treated with neoadjuvant chemoradiotherapy (CRT) were collected. Response was defined as (1) complete versus incomplete response, or (2) good (Mandard tumor regression grade (TRG) 1-2) versus poor response (TRG3-5). Prediction models were developed using combinations of the following variable groups: (1) Non-imaging: age/sex/tumor-location/tumor-morphology/CRT-surgery interval (2) Basic staging: cT-stage/cN-stage/mesorectal fascia involvement, derived from (2a) original staging reports, or (2b) expert re-evaluation (3) Advanced staging: variables from 2b combined with cTN-substaging/invasion depth/extramural vascular invasion/tumor length (4) Quantitative imaging: tumour volume + first-order histogram features (from T2W-MRI and DWI/ADC) Models were developed with data from 6 centers (n = 412) using logistic regression with the Least Absolute Shrinkage and Selector Operator (LASSO) feature selection, internally validated using repeated (n = 100) random hold-out validation, and externally validated using data from 3 centers (n = 97).

RESULTS

After external validation, the best model (including non-imaging and advanced staging variables) achieved an area under the curve of 0.60 (95%CI=0.48-0.72) to predict complete response and 0.65 (95%CI=0.53-0.76) to predict a good response. Quantitative variables did not improve model performance. Basic staging variables consistently achieved lower performance compared to advanced staging variables.

CONCLUSIONS

Overall model performance was moderate. Best results were obtained using advanced staging variables, highlighting the importance of good-quality staging according to current guidelines. Quantitative imaging features had no added value (in this heterogeneous dataset).

CLINICAL RELEVANCE STATEMENT

Predicting tumour response at baseline could aid in tailoring neoadjuvant therapies for rectal cancer. This study shows that image-based prediction models are promising, though are negatively affected by variations in staging quality and MRI acquisition, urging the need for harmonization.

KEY POINTS

This multicenter study combining clinical information and features derived from MRI rendered disappointing performance to predict response to neoadjuvant treatment in rectal cancer. Best results were obtained with the combination of clinical baseline information and state-of-the-art image-based staging variables, highlighting the importance of good quality staging according to current guidelines and staging templates. No added value was found for quantitative imaging features in this multicenter retrospective study. This is likely related to acquisition variations, which is a major problem for feature reproducibility and thus model generalizability.

Breast DWI Analyzed Before and After Gadolinium Contrast Administration-An Intrapatient Analysis on 1.5 T and 3.0 T

OBJECTIVES

Diffusion-weighted magnetic resonance imaging (MRI) is gaining popularity as an addition to standard dynamic contrast-enhanced breast MRI. Although adding diffusion-weighted imaging (DWI) to the standard protocol design would require increased scanning-time, implementation during the contrast-enhanced phase could offer a multiparametric MRI protocol without any additional scanning time. However, gadolinium within a region of interest (ROI) might affect assessments of DWI. This study aims to determine if acquiring DWI postcontrast, incorporated in an abbreviated MRI protocol, would statistically significantly affect lesion classification. In addition, the effect of postcontrast DWI on breast parenchyma was studied.

MATERIALS AND METHODS

Screening or preoperative MRIs (1.5 T/3 T) were included for this study. Diffusion-weighted imaging was acquired with single-shot spin echo-echo planar imaging before and at approximately 2 minutes after gadoterate meglumine injection. Apparent diffusion coefficients (ADCs) based on 2-dimensional ROIs of fibroglandular tissue, as well as benign and malignant lesions at 1.5 T/3.0 T, were compared with a Wilcoxon signed rank test. Diffusivity levels were compared between precontrast and postcontrast DWI with weighted κ. An overall P ≤ 0.05 was considered statistically significant.

RESULTS

No significant changes were observed in ADC mean after contrast administration in 21 patients with 37 ROI of healthy fibroglandular tissue and in the 93 patients with 93 (malignant and benign) lesions. This effect remained after stratification on B 0 . In 18% of all lesions, a diffusion level shift was observed, with an overall weighted κ of 0.75.

CONCLUSIONS

This study supports incorporating DWI at 2 minutes postcontrast when ADC is calculated based on b150-b800 with 15 mL 0.5 M gadoterate meglumine in an abbreviated multiparametric MRI protocol without requiring extra scan time.

Independent validation of CT radiomics models in colorectal liver metastases: predicting local tumour progression after ablation

OBJECTIVES

Independent internal and external validation of three previously published CT-based radiomics models to predict local tumor progression (LTP) after thermal ablation of colorectal liver metastases (CRLM).

MATERIALS AND METHODS

Patients with CRLM treated with thermal ablation were collected from two institutions to collect a new independent internal and external validation cohort. Ablation zones (AZ) were delineated on portal venous phase CT 2-8 weeks post-ablation. Radiomics features were extracted from the AZ and a 10 mm peri-ablational rim (PAR) of liver parenchyma around the AZ. Three previously published prediction models (clinical, radiomics, combined) were tested without retraining. LTP was defined as new tumor foci appearing next to the AZ up to 24 months post-ablation.

RESULTS

The internal cohort included 39 patients with 68 CRLM and the external cohort 52 patients with 78 CRLM. 34/146 CRLM developed LTP after a median follow-up of 24 months (range 5-139). The median time to LTP was 8 months (range 2-22). The combined clinical-radiomics model yielded a c-statistic of 0.47 (95%CI 0.30-0.64) in the internal cohort and 0.50 (95%CI 0.38-0.62) in the external cohort, compared to 0.78 (95%CI 0.65-0.87) in the previously published original cohort. The radiomics model yielded c-statistics of 0.46 (95%CI 0.29-0.63) and 0.39 (95%CI 0.28-0.52), and the clinical model 0.51 (95%CI 0.34-0.68) and 0.51 (95%CI 0.39-0.63) in the internal and external cohort, respectively.

CONCLUSION

The previously published results for prediction of LTP after thermal ablation of CRLM using clinical and radiomics models were not reproducible in independent internal and external validation.

CLINICAL RELEVANCE STATEMENT

Local tumour progression after thermal ablation of CRLM cannot yet be predicted with the use of CT radiomics of the ablation zone and peri-ablational rim. These results underline the importance of validation of radiomics results to test for reproducibility in independent cohorts.

KEY POINTS

• Previous research suggests CT radiomics models have the potential to predict local tumour progression after thermal ablation in colorectal liver metastases, but independent validation is lacking. • In internal and external validation, the previously published models were not able to predict local tumour progression after ablation. • Radiomics prediction models should be investigated in independent validation cohorts to check for reproducibility.

Diagnostic accuracy of CT for local staging of colon cancer: A nationwide study in the Netherlands

OBJECTIVE

To determine the accuracy of computed tomography (CT)-based staging in selecting high-risk colon cancer patients who would benefit from neoadjuvant chemotherapy while avoiding overtreatment.

METHODS

Data of adult patients diagnosed with non-metastatic primary colon cancer in 2005-2020, who underwent surgical resection without neoadjuvant chemotherapy, were retrospectively collected from the Netherlands Cancer Registry. Agreement between clinical and pathological evaluation for each T and N stage was calculated. Sensitivity and specificity analyses were conducted to predict T3-T4 and N1-N2 stages, with histopathology as the reference standard.

RESULTS

Data from 44,471 patients (median age, 71 years, 50% female) were evaluated. We included 38,915 patients with complete T stage and 39,565 patients with complete N stage for analyses. The overall clinical-pathological agreement for T stage was 59% and for N stage 57%. The sensitivity and specificity of CT to detect T3-T4 tumours were 80% (95% confidence interval (CI): 0.79, 0.80) and 76% (95% CI: 0.75, 0.77), respectively, with a positive predictive value (PPV) of 92% (95% CI: 0.92, 0.92). The sensitivity and specificity of CT to detect N1-N2 category were 62% (95% CI: 0.61, 0.63) and 70% (95% CI: 0.69, 0.71), respectively, with PPV 60% (95% CI: 0.59, 0.60).

CONCLUSION

CT-based staging shows limited accuracy in selecting colon cancer patients who would benefit from neoadjuvant therapy without risking overtreatment. Detection of lymph node metastases with CT remains unreliable.

The Prediction of Biological Features Using Magnetic Resonance Imaging in Head and Neck Squamous Cell Carcinoma: A Systematic Review and Meta-Analysis

Magnetic resonance imaging (MRI) is an indispensable, routine technique that provides morphological and functional imaging sequences. MRI can potentially capture tumor biology and allow for longitudinal evaluation of head and neck squamous cell carcinoma (HNSCC). This systematic review and meta-analysis evaluates the ability of MRI to predict tumor biology in primary HNSCC. Studies were screened, selected, and assessed for quality using appropriate tools according to the PRISMA criteria. Fifty-eight articles were analyzed, examining the relationship between (functional) MRI parameters and biological features and genetics. Most studies focused on HPV status associations, revealing that HPV-positive tumors consistently exhibited lower ADCmean (SMD: 0.82; p < 0.001) and ADCminimum (SMD: 0.56; p < 0.001) values. On average, lower ADCmean values are associated with high Ki-67 levels, linking this diffusion restriction to high cellularity. Several perfusion parameters of the vascular compartment were significantly associated with HIF-1α. Analysis of other biological factors (VEGF, EGFR, tumor cell count, p53, and MVD) yielded inconclusive results. Larger datasets with homogenous acquisition are required to develop and test radiomic-based prediction models capable of capturing different aspects of the underlying tumor biology. Overall, our study shows that rapid and non-invasive characterization of tumor biology via MRI is feasible and could enhance clinical outcome predictions and personalized patient management for HNSCC.

Coverage of Lateral Lymph Nodes in Rectal Cancer Patients with Routine Radiation Therapy Practice and Associated Locoregional Recurrence Rates

PURPOSE

Involved internal iliac and obturator lateral lymph nodes (LLNs) are a known risk factor for the occurrence of ipsilateral local recurrences (LLR) in rectal cancer. This study examined coverage of LLNs with routine radiation therapy practice in the Netherlands and associated LLR rates.

METHODS AND MATERIALS

Patients with a primary tumor ≤8 cm of the anorectal junction, cT3-4 stage, and at least 1 internal iliac or obturator LLN with short axis ≥5 mm who received neoadjuvant (chemo)radiation therapy, were selected from a national, cross-sectional study of patients with rectal cancer treated in the Netherlands in 2016. Magnetic resonance images and radiation therapy treatment plans were reviewed regarding segmented LLNs as gross tumor volume (GTV), location of LLNs within clinical target volume (CTV), and received proportion of the planned radiation therapy dose.

RESULTS

A total of 223 out of 3057 patients with at least 1 LLN ≥5 mm were selected. Of those, 180 (80.7%) LLNs were inside the CTV, of which 60 (33.3%) were segmented as GTV. Overall, 202 LLNs (90.6%) received ≥95% of the planned dose. Four-year LLR rates were not significantly higher for LLNs situated outside the CTV compared with those inside (4.0% vs 12.5%, P = .092) or when receiving <95% versus ≥95% of the planned radiation therapy dose (7.1% vs 11.3%, P = .843), respectively. Two of 7 patients who received a dose escalation of 60 Gy developed an LLR (4-year LLR rate of 28.6%).

CONCLUSIONS

This evaluation of routine radiation therapy practice showed that adequate coverage of LLNs was still associated with considerable 4-year LLR rates. Techniques resulting in better local control for patients with involved LLNs need to be explored further.

Predicting response to chemoradiotherapy in rectal cancer via visual morphologic assessment and staging on baseline MRI: a multicenter and multireader study

PURPOSE

Pre-treatment knowledge of the anticipated response of rectal tumors to neoadjuvant chemoradiotherapy (CRT) could help to further optimize the treatment. Van Griethuysen et al. proposed a visual 5-point confidence score to predict the likelihood of response on baseline MRI. Aim was to evaluate this score in a multicenter and multireader study setting and compare it to two simplified (4-point and 2-point) adaptations in terms of diagnostic performance, interobserver agreement (IOA), and reader preference.

METHODS

Twenty-two radiologists from 14 countries (5 MRI-experts,17 general/abdominal radiologists) retrospectively reviewed 90 baseline MRIs to estimate if patients would likely achieve a (near-)complete response (nCR); first using the 5-point score by van Griethuysen (1=highly unlikely to 5=highly likely to achieve nCR), second using a 4-point adaptation (with 1-point each for high-risk T-stage, obvious mesorectal fascia invasion, nodal involvement, and extramural vascular invasion), and third using a 2-point score (unlikely/likely to achieve nCR). Diagnostic performance was calculated using ROC curves and IOA using Krippendorf's alpha (α).

RESULTS

Areas under the ROC curve to predict the likelihood of a nCR were similar for the three methods (0.71-0.74). IOA was higher for the 5- and 4-point scores (α=0.55 and 0.57 versus 0.46 for the 2-point score) with best results for the MRI-experts (α=0.64-0.65). Most readers (55%) favored the 4-point score.

CONCLUSIONS

Visual morphologic assessment and staging methods can predict neoadjuvant treatment response with moderate-good performance. Compared to a previously published confidence-based scoring system, study readers preferred a simplified 4-point risk score based on high-risk T-stage, MRF involvement, nodal involvement, and EMVI.

A Deep Learning Framework with Explainability for the Prediction of Lateral Locoregional Recurrences in Rectal Cancer Patients with Suspicious Lateral Lymph Nodes

Malignant lateral lymph nodes (LLNs) in low, locally advanced rectal cancer can cause (ipsi-lateral) local recurrences ((L)LR). Accurate identification is, therefore, essential. This study explored LLN features to create an artificial intelligence prediction model, estimating the risk of (L)LR. This retrospective multicentre cohort study examined 196 patients diagnosed with rectal cancer between 2008 and 2020 from three tertiary centres in the Netherlands. Primary and restaging T2W magnetic resonance imaging and clinical features were used. Visible LLNs were segmented and used for a multi-channel convolutional neural network. A deep learning model was developed and trained for the prediction of (L)LR according to malignant LLNs. Combined imaging and clinical features resulted in AUCs of 0.78 and 0.80 for LR and LLR, respectively. The sensitivity and specificity were 85.7% and 67.6%, respectively. Class activation map explainability methods were applied and consistently identified the same high-risk regions with structural similarity indices ranging from 0.772-0.930. This model resulted in good predictive value for (L)LR rates and can form the basis of future auto-segmentation programs to assist in the identification of high-risk patients and the development of risk stratification models.

Sense and nonsense of yT-staging on MRI after chemoradiotherapy in rectal cancer

AIM

The aim of this work was to investigate the value of rectal cancer T-staging on MRI after chemoradiotherapy (ymrT-staging) in relation to the degree of fibrotic transformation of the tumour bed as assessed using the pathological tumour regression grade (pTRG) of Mandard as a standard of reference.

METHOD

Twenty two radiologists, including five rectal MRI experts and 17 'nonexperts' (general/abdominal radiologists), evaluated the ymrT stage on the restaging MRIs of 90 rectal cancer patients after chemoradiotherapy. The ymrT stage was compared with the final ypT stage at histopathology; the percentages of correct staging (ymrT = ypT), understaging (ymrT < ypT) and overstaging (ymrT > ypT) were calculated and compared between patients with predominant tumour at histopathology (pTRG4-5) and patients with predominant fibrosis (pTRG1-3). Interobserver agreement (IOA) was computed using Krippendorff's alpha.

RESULTS

Average ymrT/ypT stage concordance was 48% for the experts and 43% for the nonexperts; ymrT/ypT stage concordance was significantly higher in the pTRG4-5 subgroup (58% vs. 41% for the pTRG1-3 group; p = 0.01), with the best results for the MRI experts. Overstaging was the main source of error, especially in the pTRG1-3 subgroup (average overstaging rate 38%-44% vs. 13%-55% in the pTRG4-5 subgroup). IOA was higher for the expert versus nonexpert readers (α = 0.67 vs. α = 0.39).

CONCLUSIONS

ymrT-staging is moderately accurate; accuracy is higher in poorly responding patients with predominant tumour but low in good responders with predominant fibrosis, resulting in significant overstaging. Radiologists should shift their focus from ymrT-staging to detecting gross residual (and progressive) disease, and identifying potential candidates for organ preservation who would benefit from further clinical and endoscopic evaluation to guide final treatment planning.

Evaluation of National Surgical Practice for Lateral Lymph Nodes in Rectal Cancer in an Untrained Setting

BACKGROUND

Involved lateral lymph nodes (LLNs) have been associated with increased local recurrence (LR) and ipsi-lateral LR (LLR) rates. However, consensus regarding the indication and type of surgical treatment for suspicious LLNs is lacking. This study evaluated the surgical treatment of LLNs in an untrained setting at a national level.

METHODS

Patients who underwent additional LLN surgery were selected from a national cross-sectional cohort study regarding patients undergoing rectal cancer surgery in 69 Dutch hospitals in 2016. LLN surgery consisted of either 'node-picking' (the removal of an individual LLN) or 'partial regional node dissection' (PRND; an incomplete resection of the LLN area). For all patients with primarily enlarged (≥7 mm) LLNs, those undergoing rectal surgery with an additional LLN procedure were compared to those  undergoing only rectal resection.

RESULTS

Out of 3057 patients, 64 underwent additional LLN surgery, with 4-year LR and LLR rates of 26% and 15%, respectively. Forty-eight patients (75%) had enlarged LLNs, with corresponding recurrence rates of 26% and 19%, respectively. Node-picking (n = 40) resulted in a 20% 4-year LLR, and a 14% LLR after PRND (n = 8; p = 0.677). Multivariable analysis of 158 patients with enlarged LLNs undergoing additional LLN surgery (n = 48) or rectal resection alone (n = 110) showed no significant association of LLN surgery with 4-year LR or LLR, but suggested higher recurrence risks after LLN surgery (LR: hazard ratio [HR] 1.5, 95% confidence interval [CI] 0.7-3.2, p = 0.264; LLR: HR 1.9, 95% CI 0.2-2.5, p = 0.874).

CONCLUSION

Evaluation of Dutch practice in 2016 revealed that approximately one-third of patients with primarily enlarged LLNs underwent surgical treatment, mostly consisting of node-picking. Recurrence rates were not significantly affected by LLN surgery, but did suggest worse outcomes. Outcomes of LLN surgery after adequate training requires further research.

RadioLOGIC, a healthcare model for processing electronic health records and decision-making in breast disease

Digital health data used in diagnostics, patient care, and oncology research continue to accumulate exponentially. Most medical information, and particularly radiology results, are stored in free-text format, and the potential of these data remains untapped. In this study, a radiological repomics-driven model incorporating medical token cognition (RadioLOGIC) is proposed to extract repomics (report omics) features from unstructured electronic health records and to assess human health and predict pathological outcome via transfer learning. The average accuracy and F1-weighted score for the extraction of repomics features using RadioLOGIC are 0.934 and 0.934, respectively, and 0.906 and 0.903 for the prediction of breast imaging-reporting and data system scores. The areas under the receiver operating characteristic curve for the prediction of pathological outcome without and with transfer learning are 0.912 and 0.945, respectively. RadioLOGIC outperforms cohort models in the capability to extract features and also reveals promise for checking clinical diagnoses directly from electronic health records.

Radiomic signatures from T2W and DWI MRI are predictive of tumour hypoxia in colorectal liver metastases

BACKGROUND

Tumour hypoxia is a negative predictive and prognostic biomarker in colorectal cancer typically assessed by invasive sampling methods, which suffer from many shortcomings. This retrospective proof-of-principle study explores the potential of MRI-derived imaging markers in predicting tumour hypoxia non-invasively in patients with colorectal liver metastases (CLM).

METHODS

A single-centre cohort of 146 CLMs from 112 patients were segmented on preoperative T2-weighted (T2W) images and diffusion-weighted imaging (DWI). HIF-1 alpha immunohistochemical staining index (high/low) was used as a reference standard. Radiomic features were extracted, and machine learning approaches were implemented to predict the degree of histopathological tumour hypoxia.

RESULTS

Radiomic signatures from DWI b200 (AUC = 0.79, 95% CI 0.61-0.93, p = 0.002) and ADC (AUC = 0.72, 95% CI 0.50-0.90, p = 0.019) were significantly predictive of tumour hypoxia. Morphological T2W TE75 (AUC = 0.64, 95% CI 0.42-0.82, p = 0.092) and functional DWI b0 (AUC = 0.66, 95% CI 0.46-0.84, p = 0.069) and b800 (AUC = 0.64, 95% CI 0.44-0.82, p = 0.071) images also provided predictive information. T2W TE300 (AUC = 0.57, 95% CI 0.33-0.78, p = 0.312) and b = 10 (AUC = 0.53, 95% CI 0.33-0.74, p = 0.415) images were not predictive of tumour hypoxia.

CONCLUSIONS

T2W and DWI sequences encode information predictive of tumour hypoxia. Prospective multicentre studies could help develop and validate robust non-invasive hypoxia-detection algorithms.

CRITICAL RELEVANCE STATEMENT

Hypoxia is a negative prognostic biomarker in colorectal cancer. Hypoxia is usually assessed by invasive sampling methods. This proof-of-principle retrospective study explores the role of AI-based MRI-derived imaging biomarkers in non-invasively predicting tumour hypoxia in patients with colorectal liver metastases (CLM).

Artificial Intelligence Tool for Detection and Worklist Prioritization Reduces Time to Diagnosis of Incidental Pulmonary Embolism at CT

Purpose

To evaluate the diagnostic efficacy of artificial intelligence (AI) software in detecting incidental pulmonary embolism (IPE) at CT and shorten the time to diagnosis with use of radiologist reading worklist prioritization.

Materials and Methods

In this study with historical controls and prospective evaluation, regulatory-cleared AI software was evaluated to prioritize IPE on routine chest CT scans with intravenous contrast agent in adult oncology patients. Diagnostic accuracy metrics were calculated, and temporal end points, including detection and notification times (DNTs), were assessed during three time periods (April 2019 to September 2020): routine workflow without AI, human triage without AI, and worklist prioritization with AI.

Results

In total, 11 736 CT scans in 6447 oncology patients (mean age, 63 years ± 12 [SD]; 3367 men) were included. Prevalence of IPE was 1.3% (51 of 3837 scans), 1.4% (54 of 3920 scans), and 1.0% (38 of 3979 scans) for the respective time periods. The AI software detected 131 true-positive, 12 false-negative, 31 false-positive, and 11 559 true-negative results, achieving 91.6% sensitivity, 99.7% specificity, 99.9% negative predictive value, and 80.9% positive predictive value. During prospective evaluation, AI-based worklist prioritization reduced the median DNT for IPE-positive examinations to 87 minutes (vs routine workflow of 7714 minutes and human triage of 4973 minutes). Radiologists' missed rate of IPE was significantly reduced from 44.8% (47 of 105 scans) without AI to 2.6% (one of 38 scans) when assisted by the AI tool (P < .001).

Conclusion

AI-assisted workflow prioritization of IPE on routine CT scans in oncology patients showed high diagnostic accuracy and significantly shortened the time to diagnosis in a setting with a backlog of examinations.Keywords: CT, Computer Applications, Detection, Diagnosis, Embolism, Thorax, ThrombosisSupplemental material is available for this article.© RSNA, 2023See also the commentary by Elicker in this issue.

Predicting breast cancer types on and beyond molecular level in a multi-modal fashion

Accurately determining the molecular subtypes of breast cancer is important for the prognosis of breast cancer patients and can guide treatment selection. In this study, we develop a deep learning-based model for predicting the molecular subtypes of breast cancer directly from the diagnostic mammography and ultrasound images. Multi-modal deep learning with intra- and inter-modality attention modules (MDL-IIA) is proposed to extract important relations between mammography and ultrasound for this task. MDL-IIA leads to the best diagnostic performance compared to other cohort models in predicting 4-category molecular subtypes with Matthews correlation coefficient (MCC) of 0.837 (95% confidence interval [CI]: 0.803, 0.870). The MDL-IIA model can also discriminate between Luminal and Non-Luminal disease with an area under the receiver operating characteristic curve of 0.929 (95% CI: 0.903, 0.951). These results significantly outperform clinicians' predictions based on radiographic imaging. Beyond molecular-level test, based on gene-level ground truth, our method can bypass the inherent uncertainty from immunohistochemistry test. This work thus provides a noninvasive method to predict the molecular subtypes of breast cancer, potentially guiding treatment selection for breast cancer patients and providing decision support for clinicians.

Endoscopy and MRI for restaging early rectal cancer after neoadjuvant treatment

AIM

Chemoradiotherapy (CRT) has great potential to downstage rectal cancer. Response assessment has been investigated in locally advanced rectal cancer but not in early stage rectal cancer. The aim is to characterize the diagnostic accuracy of endoscopy performed by surgical endoscopists compared to (diffusion-weighted, DWI) MRI only and a multimodal approach combining (DWI-)MRI and endoscopic information both analysed by an abdominal radiologist for response assessment in early rectal cancer after neoadjuvant CRT.

MATERIALS AND METHODS

Patients treated with neoadjuvant CRT for early distal rectal cancer (cT1-3 N0) followed by transanal endoscopic microsurgery were included. Three separate reassessment groups were analysed for response assessment using endoscopic evaluation alone versus (DWI-)MRI alone versus the combination of endoscopy with (DWI-)MRI with a focus on sensitivity and specificity and analysis using receiver operating characteristic curves.

RESULTS

Three cohorts (N = 36, N = 25 and N = 25, respectively) were analysed for response assessment. Of the endoscopy cohort, 16 of the 36 patients had a complete response. Area under the curve was 0.69 (0.66-0.74; pooled sensitivity 55.3%, pooled specificity 80.0%). Agreement for scoring separate endoscopic features was poor to moderate. Of the (DWI-)MRI cohort, 11 of the 25 patients had a complete response. Area under the curve for (DWI-)MRI alone was 0.55 (sensitivity 72.7%, specificity 42.9%). The areas under the receiver operating characteristic curve improved to 0.68 (sensitivity 90.9%, specificity 75.0%) when (DWI-)MRI was combined with endoscopic information, with 11 out of 25 patients with a complete response. The most accurate response assessment was made by combining endoscopy and (DWI-)MRI with a high negative predictive value (90.9%).

CONCLUSION

Good and complete responders after chemoradiation of early stage rectal cancer can be best assessed using a multimodality approach combining endoscopy and (DWI-)MRI.

MRI anatomy of the rectum: key concepts important for rectal cancer staging and treatment planning

A good understanding of the MRI anatomy of the rectum and its surroundings is pivotal to ensure high-quality diagnostic evaluation and reporting of rectal cancer. With this pictorial review, we aim to provide an image-based overview of key anatomical concepts essential for treatment planning, response evaluation and post-operative assessment. These concepts include the cross-sectional anatomy of the rectal wall in relation to T-staging; differences in staging and treatment between anal and rectal cancer; landmarks used to define the upper and lower boundaries of the rectum; the anatomy of the pelvic floor and anal canal, the mesorectal fascia, peritoneum and peritoneal reflection; and guides to help discern different pelvic lymph node stations on MRI to properly stage regional and non-regional rectal lymph node metastases. Finally, this review will highlight key aspects of post-treatment anatomy, including the assessment of radiation-induced changes and the evaluation of the post-operative pelvis after different surgical resection and reconstruction techniques.

Is the generalizability of a developed artificial intelligence algorithm for COVID-19 on chest CT sufficient for clinical use? Results from the International Consortium for COVID-19 Imaging AI (ICOVAI)

OBJECTIVES

Only few published artificial intelligence (AI) studies for COVID-19 imaging have been externally validated. Assessing the generalizability of developed models is essential, especially when considering clinical implementation. We report the development of the International Consortium for COVID-19 Imaging AI (ICOVAI) model and perform independent external validation.

METHODS

The ICOVAI model was developed using multicenter data (n = 1286 CT scans) to quantify disease extent and assess COVID-19 likelihood using the COVID-19 Reporting and Data System (CO-RADS). A ResUNet model was modified to automatically delineate lung contours and infectious lung opacities on CT scans, after which a random forest predicted the CO-RADS score. After internal testing, the model was externally validated on a multicenter dataset (n = 400) by independent researchers. CO-RADS classification performance was calculated using linearly weighted Cohen's kappa and segmentation performance using Dice Similarity Coefficient (DSC).

RESULTS

Regarding internal versus external testing, segmentation performance of lung contours was equally excellent (DSC = 0.97 vs. DSC = 0.97, p = 0.97). Lung opacities segmentation performance was adequate internally (DSC = 0.76), but significantly worse on external validation (DSC = 0.59, p < 0.0001). For CO-RADS classification, agreement with radiologists on the internal set was substantial (kappa = 0.78), but significantly lower on the external set (kappa = 0.62, p < 0.0001).

CONCLUSION

In this multicenter study, a model developed for CO-RADS score prediction and quantification of COVID-19 disease extent was found to have a significant reduction in performance on independent external validation versus internal testing. The limited reproducibility of the model restricted its potential for clinical use. The study demonstrates the importance of independent external validation of AI models.

KEY POINTS

• The ICOVAI model for prediction of CO-RADS and quantification of disease extent on chest CT of COVID-19 patients was developed using a large sample of multicenter data. • There was substantial performance on internal testing; however, performance was significantly reduced on external validation, performed by independent researchers. The limited generalizability of the model restricts its potential for clinical use. • Results of AI models for COVID-19 imaging on internal tests may not generalize well to external data, demonstrating the importance of independent external validation.

Multi-modal artificial intelligence for the combination of automated 3D breast ultrasound and mammograms in a population of women with predominantly dense breasts

OBJECTIVES

To assess the stand-alone and combined performance of artificial intelligence (AI) detection systems for digital mammography (DM) and automated 3D breast ultrasound (ABUS) in detecting breast cancer in women with dense breasts.

METHODS

430 paired cases of DM and ABUS examinations from a Asian population with dense breasts were retrospectively collected. All cases were analyzed by two AI systems, one for DM exams and one for ABUS exams. A selected subset (n = 152) was read by four radiologists. The performance of AI systems was based on analysis of the area under the receiver operating characteristic curve (AUC). The maximum Youden's index and its associated sensitivity and specificity were also reported for each AI systems. Detection performance of human readers in the subcohort of the reader study was measured in terms of sensitivity and specificity.

RESULTS

The performance of the AI systems in a multi-modal setting was significantly better when the weights of AI-DM and AI-ABUS were 0.25 and 0.75, respectively, than each system individually in a single-modal setting (AUC-AI-Multimodal = 0.865; AUC-AI-DM = 0.832, p = 0.026; AUC-AI-ABUS = 0.841, p = 0.041). The maximum Youden's index for AI-Multimodal was 0.707 (sensitivity = 79.4%, specificity = 91.2%). In the subcohort that underwent human reading, the panel of four readers achieved a sensitivity of 93.2% and specificity of 32.7%. AI-multimodal achieves superior or equal sensitivity as single human readers at the same specificity operating points on the ROC curve.

CONCLUSION

Multimodal (ABUS + DM) AI systems for detecting breast cancer in women with dense breasts are a potential solution for breast screening in radiologist-scarce regions.

A deep learning-based application for COVID-19 diagnosis on CT: The Imaging COVID-19 AI initiative

BACKGROUND

Recently, artificial intelligence (AI)-based applications for chest imaging have emerged as potential tools to assist clinicians in the diagnosis and management of patients with coronavirus disease 2019 (COVID-19).

OBJECTIVES

To develop a deep learning-based clinical decision support system for automatic diagnosis of COVID-19 on chest CT scans. Secondarily, to develop a complementary segmentation tool to assess the extent of lung involvement and measure disease severity.

METHODS

The Imaging COVID-19 AI initiative was formed to conduct a retrospective multicentre cohort study including 20 institutions from seven different European countries. Patients with suspected or known COVID-19 who underwent a chest CT were included. The dataset was split on the institution-level to allow external evaluation. Data annotation was performed by 34 radiologists/radiology residents and included quality control measures. A multi-class classification model was created using a custom 3D convolutional neural network. For the segmentation task, a UNET-like architecture with a backbone Residual Network (ResNet-34) was selected.

RESULTS

A total of 2,802 CT scans were included (2,667 unique patients, mean [standard deviation] age = 64.6 [16.2] years, male/female ratio 1.3:1). The distribution of classes (COVID-19/Other type of pulmonary infection/No imaging signs of infection) was 1,490 (53.2%), 402 (14.3%), and 910 (32.5%), respectively. On the external test dataset, the diagnostic multiclassification model yielded high micro-average and macro-average AUC values (0.93 and 0.91, respectively). The model provided the likelihood of COVID-19 vs other cases with a sensitivity of 87% and a specificity of 94%. The segmentation performance was moderate with Dice similarity coefficient (DSC) of 0.59. An imaging analysis pipeline was developed that returned a quantitative report to the user.

CONCLUSION

We developed a deep learning-based clinical decision support system that could become an efficient concurrent reading tool to assist clinicians, utilising a newly created European dataset including more than 2,800 CT scans.

Comparison of MRI response evaluation methods in rectal cancer: a multicentre and multireader validation study

OBJECTIVES

To compare four previously published methods for rectal tumor response evaluation after chemoradiotherapy on MRI.

METHODS

Twenty-two radiologists (5 rectal MRI experts, 17 general/abdominal radiologists) retrospectively reviewed the post-chemoradiotherapy MRIs of 90 patients, scanned at 10 centers (with non-standardized protocols). They applied four response methods; two based on T2W-MRI only (MRI tumor regression grade (mrTRG); split-scar sign), and two based on T2W-MRI+DWI (modified-mrTRG; DWI-patterns). Image quality was graded using a 0-6-point score (including slice thickness and in-plane resolution; sequence angulation; DWI b-values, signal-to-noise, and artefacts); scores < 4 were classified below average. Mixed model linear regression was used to calculate average sensitivity/specificity/accuracy to predict a complete response (versus residual tumor) and assess the impact of reader experience and image quality. Group interobserver agreement (IOA) was calculated using Krippendorff's alpha. Readers were asked to indicate their preferred scoring method(s).

RESULTS

Average sensitivity/specificity/accuracy was 57%/64%/62% (mrTRG), 36%/79%/66% (split-scar), 40%/79%/67% (modified-mrTRG), and 37%/82%/68% (DWI-patterns); mrTRG showed higher sensitivity but lower specificity and accuracy (p < 0.001) compared to the other methods. IOA was lower for the split scar method (0.18 vs. 0.39-0.43). Higher reader experience had a significant positive effect on diagnostic performance and IOA (except for the split scar sign); below-average imaging quality had a significant negative effect on diagnostic performance. DWI pattern was selected as the preferred method by 73% of readers.

CONCLUSIONS

Methods incorporating DWI showed the most favorable results when combining diagnostic performance, IOA, and reader preference. Reader experience and image quality clearly impacted diagnostic performance emphasizing the need for state-of-the-art imaging and dedicated radiologist training.

KEY POINTS

• In a multireader study comparing 4 MRI methods for rectal tumor response evaluation, those incorporating DWI showed the best results when combining diagnostic performance, IOA, and reader preference. • The most preferred method (by 73% of readers) was the "DWI patterns" approach with an accuracy of 68%, high specificity of 82%, and group IOA of 0.43. • Reader experience level and MRI quality had an evident effect on diagnostic performance and IOA.

Federated learning enables big data for rare cancer boundary detection

Although machine learning (ML) has shown promise across disciplines, out-of-sample generalizability is concerning. This is currently addressed by sharing multi-site data, but such centralization is challenging/infeasible to scale due to various limitations. Federated ML (FL) provides an alternative paradigm for accurate and generalizable ML, by only sharing numerical model updates. Here we present the largest FL study to-date, involving data from 71 sites across 6 continents, to generate an automatic tumor boundary detector for the rare disease of glioblastoma, reporting the largest such dataset in the literature (n = 6, 314). We demonstrate a 33% delineation improvement for the surgically targetable tumor, and 23% for the complete tumor extent, over a publicly trained model. We anticipate our study to: 1) enable more healthcare studies informed by large diverse data, ensuring meaningful results for rare diseases and underrepresented populations, 2) facilitate further analyses for glioblastoma by releasing our consensus model, and 3) demonstrate the FL effectiveness at such scale and task-complexity as a paradigm shift for multi-site collaborations, alleviating the need for data-sharing.

Quantification and reduction of cross-vendor variation in multicenter DWI MR imaging: results of the Cancer Core Europe imaging task force

OBJECTIVES

In the Cancer Core Europe Consortium (CCE), standardized biomarkers are required for therapy monitoring oncologic multicenter clinical trials. Multiparametric functional MRI and particularly diffusion-weighted MRI offer evident advantages for noninvasive characterization of tumor viability compared to CT and RECIST. A quantification of the inter- and intraindividual variation occurring in this setting using different hardware is missing. In this study, the MRI protocol including DWI was standardized and the residual variability of measurement parameters quantified.

METHODS

Phantom and volunteer measurements (single-shot T2w and DW-EPI) were performed at the seven CCE sites using the MR hardware produced by three different vendors. Repeated measurements were performed at the sites and across the sites including a traveling volunteer, comparing qualitative and quantitative ROI-based results including an explorative radiomics analysis.

RESULTS

For DWI/ADC phantom measurements using a central post-processing algorithm, the maximum deviation could be decreased to 2%. However, there is no significant difference compared to a decentralized ADC value calculation at the respective MRI devices. In volunteers, the measurement variation in 2 repeated scans did not exceed 11% for ADC and is below 20% for single-shot T2w in systematic liver ROIs. The measurement variation between sites amounted to 20% for ADC and < 25% for single-shot T2w. Explorative radiomics classification experiments yield better results for ADC than for single-shot T2w.

CONCLUSION

Harmonization of MR acquisition and post-processing parameters results in acceptable standard deviations for MR/DW imaging. MRI could be the tool in oncologic multicenter trials to overcome the limitations of RECIST-based response evaluation.

KEY POINTS

• Harmonizing acquisition parameters and post-processing homogenization, standardized protocols result in acceptable standard deviations for multicenter MR-DWI studies. • Total measurement variation does not to exceed 11% for ADC in repeated measurements in repeated MR acquisitions, and below 20% for an identical volunteer travelling between sites. • Radiomic classification experiments were able to identify stable features allowing for reliable discrimination of different physiological tissue samples, even when using heterogeneous imaging data.

Peroperative extent of peritoneal metastases affects the surgical outcome and survival in advanced ovarian cancer

OBJECTIVE

Determining whether cytoreductive surgery (CRS) is feasible in patients with advanced ovarian cancer and whether extensive surgery is justified is challenging. Accurate patient selection for CRS based on pre- and peroperative parameters will be valuable. The aim of this study is to assess the association between the extent of peritoneal metastases as determined during surgery and completeness of interval CRS and survival.

METHODS

This single-center observational cohort study included consecutive patients with newly diagnosed stage III-IV epithelial ovarian cancer who received neoadjuvant chemotherapy and underwent interval CRS. The 7 Region Count (7RC) was recorded during surgical exploration to systematically quantify the extent of peritoneal metastases. Logistic regression analysis was performed to predict surgical outcomes, and Cox regression analysis was done for survival outcomes.

RESULTS

A total of 316 patients were included for analyses. The median 7RC was 4 (interquartile range: 2-6). Complete CRS was performed in 58%, optimal CRS in 30%, and incomplete CRS in 12% of patients. A higher 7RC was independently associated with lower odds of complete or optimal CRS in multivariable analysis (odds ratio [OR] = 0.45, 95% confidence interval [CI]: 0.33-0.63, p < 0.001). Similarly, a higher 7RC was independently associated with worse progression-free survival (hazard ratio [HR] = 1.17, 95% CI 1.08-1.26, p < 0.001) and overall survival (HR = 1.14, 95% CI 1.04-1.25, p = 0.007).

CONCLUSION

The extent of peritoneal metastases, as expressed by the 7RC during surgery, is an independent predictor for completeness of CRS and has independent prognostic value for progression-free survival and overall survival in addition to completeness of CRS.

Retrospective evaluation of national MRI reporting quality for lateral lymph nodes in rectal cancer patients and concordance with prospective re-evaluation following additional training

Objectives

The presence and size of lateral lymph nodes (LLNs) are important factors influencing treatment decisions for rectal cancer. Awareness of the clinical relevance and describing LLNs in MRI reports is therefore essential. This study assessed whether LLNs were mentioned in primary MRI reports at a national level and investigated the concordance with standardised re-review.

Methods

This national, retrospective, cross-sectional cohort study included 1096 patients from 60 hospitals treated in 2016 for primary cT3-4 rectal cancer ≤ 8 cm from the anorectal junction. Abdominal radiologists re-reviewed all MR images following a 2-h training regarding LLNs.

Results

Re-review of MR images identified that 41.0% of enlarged (≥ 7 mm) LLNs were not mentioned in primary MRI reports. A contradictory anatomical location was stated for 73.2% of all LLNs and a different size (≥/< 7 mm) for 41.7%. In total, 49.4% of  all cases did not mention LLNs in primary MRI reports. Reporting LLNs was associated with stage (cT3N0 44.3%, T3N+/T4 52.8%, p = 0.013), cN stage (N0 44.1%, N1 48.6%, N2 59.5%, p < 0.001), hospital type (non-teaching 34.6%, teaching 52.2%, academic 53.2% p = 0.006) and annual rectal cancer resection volumes (low 34.8%, medium 47.7%, high 57.3% p < 0.001). For LLNs present according to original MRI reports (n = 226), 64.2% also mentioned a short-axis size, 52.7% an anatomical location and 25.2% whether it was deemed suspicious.

Conclusions

Almost half of the primary MRI reports for rectal cancer patients treated in the Netherlands in 2016 did not mention LLNs. A significant portion of enlarged LLNs identified during re-review were also not mentioned originally, with considerable discrepancies for location and size. These results imply insufficient awareness and indicate the need for templates, education and training.

GEP-NET radiomics: a systematic review and radiomics quality score assessment

OBJECTIVE

The number of radiomics studies in gastroenteropancreatic neuroendocrine tumours (GEP-NETs) is rapidly increasing. This systematic review aims to provide an overview of the available evidence of radiomics for clinical outcome measures in GEP-NETs, to understand which applications hold the most promise and which areas lack evidence.

METHODS

PubMed, Embase, and Wiley/Cochrane Library databases were searched and a forward and backward reference check of the identified studies was executed. Inclusion criteria were (1) patients with GEP-NETs and (2) radiomics analysis on CT, MRI or PET. Two reviewers independently agreed on eligibility and assessed methodological quality with the radiomics quality score (RQS) and extracted outcome data.

RESULTS

In total, 1364 unique studies were identified and 45 were included for analysis. Most studies focused on GEP-NET grade and differential diagnosis of GEP-NETs from other neoplasms, while only a minority analysed treatment response or long-term outcomes. Several studies were able to predict tumour grade or to differentiate GEP-NETs from other lesions with a good performance (AUCs 0.74-0.96 and AUCs 0.80-0.99, respectively). Only one study developed a model to predict recurrence in pancreas NETs (AUC 0.77). The included studies reached a mean RQS of 18%.

CONCLUSION

Although radiomics for GEP-NETs is still a relatively new area, some promising models have been developed. Future research should focus on developing robust models for clinically relevant aims such as prediction of response or long-term outcome in GEP-NET, since evidence for these aims is still scarce.

KEY POINTS

• The majority of radiomics studies in gastroenteropancreatic neuroendocrine tumours is of low quality. • Most evidence for radiomics is available for the identification of tumour grade or differentiation of gastroenteropancreatic neuroendocrine tumours from other neoplasms. • Radiomics for the prediction of response or long-term outcome in gastroenteropancreatic neuroendocrine tumours warrants further research.