Predicting Response to Total Neoadjuvant Treatment (TNT) in Locally Advanced Rectal Cancer Based on Multiparametric Magnetic Resonance Imaging: A Retrospective Study

Machine learning in predicting pathological complete response to neoadjuvant chemoradiotherapy in rectal cancer using MRI: a systematic review and meta-analysis

OBJECTIVES

To evaluate the performance of machine learning models in predicting pathological complete response (pCR) to neoadjuvant chemoradiotherapy (nCRT) in patients with rectal cancer using magnetic resonance imaging.

METHODS

We searched PubMed, Embase, Cochrane Library, and Web of Science for studies published before March 2024. The Quality Assessment of Diagnostic Accuracy Studies 2 (QUADAS-2) was used to assess the methodological quality of the included studies, random-effects models were used to calculate sensitivity and specificity, I2 values were used for heterogeneity measurements, and subgroup analyses were carried out to detect potential sources of heterogeneity.

RESULTS

A total of 1699 patients from 24 studies were included. For machine learning models in predicting pCR to nCRT, the meta-analysis calculated a pooled area under the curve (AUC) of 0.91 (95% CI, 0.88-0.93), pooled sensitivity of 0.83 (95% CI, 0.74-0.89), and pooled specificity of 0.86 (95% CI, 0.80-0.91). We investigated 6 studies that mainly contributed to heterogeneity. After performing meta-analysis again excluding these 6 studies, the heterogeneity was significantly reduced. In subgroup analysis, the pooled AUC of the deep-learning model was 0.93 and 0.89 for the traditional statistical model; the pooled AUC of studies that used diffusion-weighted imaging (DWI) was 0.90 and 0.92 in studies that did not use DWI; the pooled AUC of studies conducted in China was 0.93, and was 0.83 in studies conducted in other countries.

CONCLUSIONS

This systematic study showed that machine learning has promising potential in predicting pCR to nCRT in patients with locally advanced rectal cancer. Compared to traditional machine learning models, although deep-learning-based studies are less predominant and more heterogeneous, they are able to obtain higher AUC.

ADVANCES IN KNOWLEDGE

Compared to traditional machine learning models, deep-learning-based studies are able to obtain higher AUC, although they are less predominant and more heterogeneous. Together with clinical information, machine learning-based models may bring us closer towards precision medicine.

Oncological outcomes after a pathological complete response following total neoadjuvant therapy or chemoradiotherapy for high-risk locally advanced rectal cancer in the RAPIDO trial

BACKGROUND

A pathological complete response (pCR) following chemoradiation (CRT) or short-course radiotherapy (scRT) leads to a favourable prognosis in patients with rectal cancer. Total neo-adjuvant therapy (TNT) doubles the pCR rate, but it is unknown whether oncological outcomes remain favourable and whether the same characteristics are associated with pCR as after CRT.

METHODS

Comparison between patients with pCR in the RAPIDO trial in the experimental [EXP] (scRT, chemotherapy, surgery, as TNT) and standard-of-care treatment [STD] (CRT, surgery, postoperative chemotherapy depending on hospital policy) groups. Primary and secondary outcomes were time-to-recurrence (TTR), overall survival (OS) and association between patient, tumour, and treatment characteristics and pCR.

RESULTS

Among patients with a resection within six months after preoperative treatment, 120/423 (28%) [EXP] and 57/398 (14%) [STD] achieved a pCR. Following pCR, 5-year cumulative TTR and OS rates in the EXP and STD arms were 8% vs. 7% (hazard ratio 1.04, 95%CI 0.32-3.38) and 94% vs. 93% (hazard ratio 1.41, 95%CI 0.51-3.92), respectively. Besides the EXP treatment (odds ratio 2.70, 95%CI 1.83-3.97), pre-treatment carcinoembryonic antigen (CEA) <5, pre-treatment tumour size <40 mm and cT2 were associated with pCR. Distance from the anal verge was the only characteristic with a statistically significant difference in association with pCR between the EXP and STD treatment (Pinteraction=0.042). pCR rates did not increase with prolonged treatment time.

CONCLUSIONS

The doubled pCR rate of TNT compared to CRT results in similar oncological outcomes. Characteristics associated with pCR are the EXP treatment, normal CEA, and small tumour size.

Predictive role of diffusion-weighted MRI in the assessment of response to total neoadjuvant therapy in locally advanced rectal cancer

OBJECTIVE

To investigate the predictive role of diffusion-weighted magnetic resonance imaging (DW-MRI) in the assessment of response to total neoadjuvant therapy (TNT) in patients with locally advanced rectal cancer (LARC).

METHODS

In this single-center retrospective study, patients with LARC who underwent staging MRI and TNT were enrolled. MRI-based staging, tumor volume, and DWI-ADC values were analyzed. Patients were classified as complete responders (pCR) and non-complete responders (non-pCR), according to post-surgical outcome. Pre-treatment ADC values were compared to pathological outcome, post-treatment downstaging, and reduction of tumor volume. The diagnostic accuracy of DWI-ADC in differentiating between pCR and non-pCR groups was calculated with receiver operating characteristic (ROC) analysis.

RESULTS

A total of 36 patients were evaluated (pCR, n = 20; non-pCR, n = 16). Pre-treatment ADC values were significantly different between the two groups (p = 0.034), while no association was found between pre-TNT tumor volume and pathological response. ADC values showed significant correlations with loco-regional downstaging after therapy (r = -0.537, p = 0.022), and with the reduction of tumor volume (r = -0.480, p = 0.044). ADC values were able to differentiate pCR from non-pCR patients with a sensitivity of 75% and specificity of 70%.

CONCLUSIONS

ADC values on pre-treatment MRI were strongly associated with the outcome in patients with LARC, both in terms of pathological response and in loco-regional downstaging after TNT, suggesting the use of DW-MRI as a potential predictive tool of response to therapy.

KEY POINTS

• ADC values of pre-TNT MRI examinations of patients with LARC were significantly associated with a pathological complete response (pCR) and with post-treatment regression of TNM staging. • An ADC value of 1.042 ×10^-3 mm²/s was found to be the optimal cutoff value for discriminating between pCR and non-pCR patients, with a sensitivity of 75% and specificity of 70%. • DW-MRI proved to have a potential predictive role in the assessment of response to therapy in patients with LARC, throughout the analysis of ADC map values.

Clinical outcome of neoadjuvant chemoradiation in rectal cancer treatment

To determine the clinical and pathological outcome of locally advanced rectal cancer patients treated with neoadjuvant chemoradiation (chemoradiotherapy [CRT]) followed by curative surgery and to identify predictive factors of pathological complete response (pCR).Locally advanced rectal cancer patients undergoing CRT followed by curative surgery from January 2012 to December 2017 were included. Patient's demographic data, pretreatment tumor characteristics, type of CRT regimens, type of surgery, postoperative complications, pathological reports and follow up records were analyzed. Univariate and multivariate analyses were applied to identify predictive factors for pCR. Five-year disease free and overall survival were estimated by Kaplan-Meier method and compared between pCR and non-pCR groups.A total of 85 patients were analyzed. Eighteen patients (21.1%) achieved pCR. The sphincter-saving surgery rate was 57.6%. After univariate analyses, tumor length >4 cm (P = .007) and positive lymph nodes (P = .040) were significantly associated with decreased rate of pCR. Complete clinical response was significantly associated with higher rate of pCR (P = .015). Multivariate analyses demonstrated that tumor length >4 cm (P = .010) was significantly associated with decreased rate of pCR. After a median follow-up of 65 months (IQR 34-79), the calculated 5-year overall survival and disease-free survival rates were 81.4% and 69.7%, respectively. Patients who achieved pCR tend to had longer 5-year disease-free survival (P = .355) and overall survival (P = .361) than those who did not.Tumor length >4 cm was associated with decreased rate of pCR in locally advanced rectal cancer who had CRT followed by surgery. Longer waiting time or more intense adjuvant treatment may be considered to improved pCR and oncological outcomes.