Diagnostic accuracy and pitfalls of MRI for restaging locally advanced rectal cancer in patients following anti-PD1 therapy plus neoadjuvant chemoradiotherapy: a multicenter study.

PURPOSE

Effectiveness of programmed cell death 1 blockade (anti-PD1) treatment plus neoadjuvant chemoradiotherapy (NCRT) in patients with locally advanced rectal cancer and proficient mismatch repair (pMMR-LARC) has been recently proven. However, the role of MRI in tumor restaging following anti-PD1 plus NCRT is less established. This study aims to evaluate the diagnostic performance and challenges of MRI for restaging pMMR-LARC patients after anti-PD1 plus NCRT treatment.

METHODS

This prospective multicenter study involved pMMR-LARC patients exhibiting following anti-PD1 plus NCRT treatment from 2021 to 2022. The evaluation of preoperative treatment response included assessments of complete response and near complete response (CR/nCR), N category, extramural venous invasion (EMVI) and mesorectal fascia (MRF), all of which were analyzed using MRI. The diagnostic accuracy was assessed using pathology as the reference standard.

RESULTS

A cohort of 44 patients (mean age, 60.5 years ± 10.4 [SD]; 15 females) was evaluated. The pathologic CR/nCR rate was 75.0%. MRI demonstrated an accuracy of 86.4% for identifying pathologic CR/nCR (pathologic tumor regression grade 0-1). The accuracy for involvement of ypN, circumferential resection margin (CRM) and ypEMVI were 86.4%, 100% and 90.9%, respectively. Minimal fibrosis (p = 0.022) and irregular fibrosis (p = 0.030) rather than oedema or mucin, contributed to misinterpretations in tumor treatment response assessment and ypEMVI using MRI, respectively.

CONCLUSION

Standard MRI restaging criteria was suitable for tumor restaging in pMMR-LARC patients following anti-PD1 plus NCRT therapy. Fibrotic-pattern-based interpretation methodology are needed in improving diagnostic accuracy of MRI.