Assessment of Clinical Complete Response After Chemoradiation for Rectal Cancer with Digital Rectal Examination, Endoscopy, and MRI: Selection for Organ-Saving Treatment

Abstract

Background: The response to chemoradiotherapy (CRT) for rectal cancer can be assessed by clinical examination, consisting of digital rectal examination (DRE) and endoscopy, and by MRI. A high accuracy is required to select complete response (CR) for organ-preserving treatment. The aim of this study was to evaluate the value of clinical examination (endoscopy with or without biopsy and DRE), T2W-MRI, and diffusion-weighted MRI (DWI) for the detection of CR after CRT.

Methods: This prospective cohort study in a university hospital recruited 50 patients who underwent clinical assessment (DRE, endoscopy with or without biopsy), T2W-MRI, and DWI at 6-8 weeks after CRT. Confidence levels were used to score the likelihood of CR. The reference standard was histopathology or recurrence-free interval of >12 months in cases of wait-and-see approaches. Diagnostic performance was calculated by area under the receiver operator characteristics curve, with corresponding sensitivities and specificities. Strategies were assessed and compared by use of likelihood ratios.

Results: Seventeen (34 %) of 50 patients had a CR. Areas under the curve were 0.88 (0.78-1.00) for clinical assessment and 0.79 (0.66-0.92) for T2W-MRI and DWI. Combining the modalities led to a posttest probability for predicting a CR of 98 %. Conversely, when all modalities indicated residual tumor, 15 % of patients still experienced CR.

Conclusions: Clinical assessment after CRT is the single most accurate modality for identification of CR after CRT. Addition of MRI with DWI further improves the diagnostic performance, and the combination can be recommended as the optimal strategy for a safe and accurate selection of CR after CRT.

Figures

Fig. 1

Response assessment with T2W-MRI (a–c) and with endoscopy (d–f). Pre- and post-CRT MR images are shown. T indicates tumor; arrows indicate scar or residual tumor after CRT. a Typical CR at T2W-MRI, b equivocal image at T2W-MRI, and c obvious residual tumor at T2W-MRI. d Typical endoluminal image of CR with white scar with teleangiectasia. e Small ulcer with smooth edges (arrows) but without residual polypoid tissue. Patients imaged in (d) and (e) experienced sustained clinical CR at follow-up. f Example of large ulcer that was deemed residual tumor after CRT

Fig. 2

Example of patient with a CR where T2W-MRI (a) revealed marked hypointense residual wall thickening resulting with an equivocal (confidence level 2) score. Clinical assessment (b) revealed a white scar with some stenosis and distortion, and small superficial ulceration, also resulting in an equivocal score. DWI (c) revealed absence of diffusion restriction indicating CR

Fig. 3

ROC curves for modalities. Clinical assessment consists of endoscopy, DRE, and biopsy result (if available)

Fig. 4

a Tumor (asterisks) before CRT. After CRT at T2W-MRI (b), fibrosis (arrows) is found with absence of high signal on DWI (c), suggestive of a CR. At endoscopy (d), a residual ulcer (arrows) is found, indicating residual tumor. Patient refused surgery and has been followed up for 3.5 years with stable MR image and a healed ulcer (e, arrows), so is classified as having experienced CR

Fig. 5

a, b Distal tumor (asterisks) before CRT at T2W-MRI and c DWI. After CRT at T2W-MRI (d) and DWI (e), residual tumor was suspected (arrows). At endoscopy (f), CR (arrows) was determined, and the patient was treated with wait-and-see policy. After 3 months, DWI became normal; patient remained free of recurrent disease at 3.8 years of follow-up