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

Abstract

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.

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

“…With only the finest feature map (stage 1), the AUC of external validation could reach up to 0.79 (0.70−0.87) using only DWI. For CR prediction, MLNet achieved AUC of 0.66 (0.55−0.77), outperforming the current state-of-the-art by Schurink et al 28 on the same external cohort. Schurink et al 28 developed a clinical-imaging model to predict CR.…”

“…For CR prediction, MLNet achieved AUC of 0.66 (0.55−0.77), outperforming the current state-of-the-art by Schurink et al 28 on the same external cohort. Schurink et al 28 developed a clinical-imaging model to predict CR. The best-performing model, using non-imaging (weeks to surgery) and advanced staging variables (tumour height, T and N staging, invasion depth and tumour length), achieved an AUC of 0.60 (0.53−0.76).…”

“…We used a dataset collected as part of a previously published multi-centre study, which included the baseline staging MRI (DWI and T2W) of 509 patient cases (obtained from one university hospital, seven large teaching hospitals and one comprehensive cancer centres from Southeren and Northern part of the Netherlands) with locally advanced rectal cancer undergoing neoadjuvant CRT. Further in- and exclusion criteria were according to those described by Schurink et al 28 . Baseline T and N staging variables cT-stage (cT1-2, cT3, cT4), cN-stage (cN0, cN1, cN2) were derived from the original staging reports that were performed by a multitude of readers.…”

“…The best-performing model, using non-imaging (weeks to surgery) and advanced staging variables (tumour height, T and N staging, invasion depth and tumour length), achieved an AUC of 0.60 (0.53−0.76). Like other Radiomics-based or tumour-centre crop-based models, one limitation of the study from Schurink et al 28 was that manually annotated segmentation for both development and test cohorts by experienced radiologists was required. To solve this, Jin et al 23 have proposed a multi-tasking learning model with pre-and-post multiparametric MRI for both segmentation and pCR assessment and the model shows the state-of-the-art results with single-centre data.…”

“…There are some limitations of the study. First of all, despite MLNet outperforming the current state-of-the-art 28 , the sensitivity (0.61) and positive predictive value (PPV) (0.44) were comparatively low, indicating that MLNet’s ability to correctly identify responders was limited, which hinders the implementation of the pipeline in the clinical workflow. The relatively low response rate (around 30%) could be one of the contributors to low sensitivity and PPV.…”