Reporting of patient characteristics and stratification factors in phase 3 trials investigating first-line systemic treatment of metastatic colorectal cancer: A systematic review

Predicting rectal cancer prognosis from histopathological images and clinical information using multi-modal deep learning

Objective

The objective of this study was to provide a multi-modal deep learning framework for forecasting the survival of rectal cancer patients by utilizing both digital pathological images data and non-imaging clinical data.

Materials and methods

The research included patients diagnosed with rectal cancer by pathological confirmation from January 2015 to December 2016. Patients were allocated to training and testing sets in a randomized manner, with a ratio of 4:1. The tissue microarrays (TMAs) and clinical indicators were obtained. Subsequently, we selected distinct deep learning models to individually forecast patient survival. We conducted a scanning procedure on the TMAs in order to transform them into digital pathology pictures. Additionally, we performed pre-processing on the clinical data of the patients. Subsequently, we selected distinct deep learning algorithms to conduct survival prediction analysis using patients' pathological images and clinical data, respectively.

Results

A total of 292 patients with rectal cancer were randomly allocated into two groups: a training set consisting of 234 cases, and a testing set consisting of 58 instances. Initially, we make direct predictions about the survival status by using pre-processed Hematoxylin and Eosin (H&E) pathological images of rectal cancer. We utilized the ResNest model to extract data from histopathological images of patients, resulting in a survival status prediction with an AUC (Area Under the Curve) of 0.797. Furthermore, we employ a multi-head attention fusion (MHAF) model to combine image features and clinical features in order to accurately forecast the survival rate of rectal cancer patients. The findings of our experiment show that the multi-modal structure works better than directly predicting from histopathological images. It achieves an AUC of 0.837 in predicting overall survival (OS).

Conclusions

Our study highlights the potential of multi-modal deep learning models in predicting survival status from histopathological images and clinical information, thus offering valuable insights for clinical applications.

Association of COVID-19 Lockdown With the Tumor Burden in Patients With Newly Diagnosed Metastatic Colorectal Cancer

Importance

The COVID-19 pandemic has been associated with substantial reduction in screening, case identification, and hospital referrals among patients with cancer. However, no study has quantitatively examined the implications of this correlation for cancer patient management.

Objective

To evaluate the association of the COVID-19 pandemic lockdown with the tumor burden of patients who were diagnosed with metastatic colorectal cancer (mCRC) before vs after lockdown.

Design, Setting, and Participants

This cohort study analyzed participants in the screening procedure of the PANIRINOX (Phase II Randomized Study Comparing FOLFIRINOX + Panitumumab vs FOLFOX + Panitumumab in Metastatic Colorectal Cancer Patients Stratified by RAS Status from Circulating DNA Analysis) phase 2 randomized clinical trial. These newly diagnosed patients received care at 1 of 18 different clinical centers in France and were recruited before or after the lockdown was enacted in France in the spring of 2020. Patients underwent a blood-sampling screening procedure to identify their RAS and BRAF tumor status.

Exposures

mCRC.

Main Outcomes and Measures

Circulating tumor DNA (ctDNA) analysis was used to identify RAS and BRAF status. Tumor burden was evaluated by the total plasma ctDNA concentration. The median ctDNA concentration was compared in patients who underwent screening before (November 11, 2019, to March 9, 2020) vs after (May 14 to September 3, 2020) lockdown and in patients who were included from the start of the PANIRINOX study.

Results

A total of 80 patients were included, of whom 40 underwent screening before and 40 others underwent screening after the first COVID-19 lockdown in France. These patients included 48 men (60.0%) and 32 women (40.0%) and had a median (range) age of 62 (37-77) years. The median ctDNA concentration was statistically higher in patients who were newly diagnosed after lockdown compared with those who were diagnosed before lockdown (119.2 ng/mL vs 17.3 ng/mL; P < .001). Patients with mCRC and high ctDNA concentration had lower median survival compared with those with lower concentration (14.7 [95% CI, 8.8-18.0] months vs 20.0 [95% CI, 14.1-32.0] months). This finding points to the potential adverse consequences of the COVID-19 pandemic and related lockdown.

Conclusions and Relevance

This cohort study found that tumor burden differed between patients who received an mCRC diagnosis before vs after the first COVID-19 lockdown in France. The findings of this study suggest that CRC is a major area for intervention to minimize pandemic-associated delays in screening, diagnosis, and treatment.