Whole-body diffusion-weighted MRI for operability assessment in patients with colorectal cancer and peritoneal metastases

Abstract

Background: Correct staging of patients with colorectal cancer is of utmost importance for the prediction of operability. Although computed tomography (CT) has a good overall performance, estimation of peritoneal cancer spread is a known weakness, a problem that cannot always be overcome by Fluorine-18 fluorodeoxyglucose positron emission tomography/computed tomography (18F-FDG-PET/CT); especially in infiltrative and miliary disease spread. Due to its high spatial and contrast resolution magnetic resonance imaging (MRI) with diffusion-weighted imaging (DWI) might have a better performance. Our aim was to evaluate the added value of whole-body diffusion-weighted MRI (WB-DWI/MRI) to CT for prediction of peritoneal cancer spread and operability assessment in colorectal cancer patients with clinically suspected peritoneal carcinomatosis (PC).

Methods: This institutional review board approved retrospective study included sixty colorectal cancer patients who underwent WB-DWI/MRI in addition to CT for clinically suspected peritoneal metastases. WB-DWI/MRI and CT were assessed for detecting PC following the peritoneal cancer index (PCI), determination of PCI-score categorized as PC < 12, PCI = 12-15 and PCI > 15, detection of nodal and distant metastases and estimation of overall operability. Histopathology after surgery and biopsy and/or 6 months follow-up were used as reference standard.

Results: For detection of PC, CT had 43.2% sensitivity, 95.6% specificity, 84.5% positive predictive value (PPV) and 75.2% negative predictive value (NPV). WB-DWI/MRI had 97.8% sensitivity, 93.2% specificity, 88.9% PPV and 98.7% NPV. WB-DWI/MRI enabled better detection of inoperable distant metastases (all 12 patients) than CT (2/12 patients) and significantly improved prediction of PCI category [WB-DWI/MRI PCI < 12: 37/39 patients (94.9%); PCI = 12-15: 4/4 patients (100%); PCI > 15: 16/17 patients (94.1%) versus CT PCI < 12: 38/39 patients (97.4%); PCI = 12-15: 0/4 patients (0%); PCI > 15: 2/17 patients (11.8%); p < 0.0001)]. WB-DWI/MRI improved prediction of inoperability over CT with 90.6% sensitivity compared to 25% (p < 0.0001).

Conclusions: WB-DWI/MRI significantly outperformed CT for estimation of spread of PC, overall staging and prediction of operability. Pending validation in larger prospective trials, WB-DWI/MRI could be used to guide surgical planning and minimize unnecessary exploratory laparotomies.

Keywords: CRS-HIPEC; CT; Colorectal peritoneal carcinomatosis; FDG-PET/CT; Metastases; Operability assessment; PCI; WB-DWI/MRI.

Conflict of interest statement

Ethics approval and consent to participate

This retrospective study was approved by the institutional review board (Ethische Commissie, S60271). Informed consent was waived.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Figures

Fig. 1

WB-DWI/MRI in a 56 year old male patient. On image A 9 of the 13 regions of the peritoneal cancer index (PCI) are displayed (regions 0–8). Region 9 is the proximal jejunum, region 10 the distal jejunum, region 11 the proximal ileum and region 12 the distal ileum (not displayed on the image). This patient was diagnosed with a caecal adenocarcinoma with extensive peritoneal metastases, for which he received chemotherapy during 5 months. Afterwards, a WB-DWI/MRI was performed to evaluate operability. On the coronal b1000 DWI (a), with a zoomed-in image (b) and the correlating coronal post-contrast T1-weighted image (c), good therapy response was seen with residual disease on the small bowel wall (arrowheads in c) with accompanying limited nodules with diffusion restriction in the small bowel mesentery (arrowheads in b). The patient was given the benefit of the doubt and underwent surgery. Surgery, however, revealed diffuse serosal metastases in the small bowel, limiting curative resection. Instead, optimal palliative care was provided. This patient was scored as a false negative interpretation of the WB-DWI/MRI

Fig. 2

Coronal CT (a) and WB-DWI/MRI b1000 (b) image of a 73 year old male patient with a primary tumour in the descending colon and synchronous peritoneal carcinomatosis. CT images estimated the PCI to be 7 (partially shown with the arrowheads in A as a hypodense thickening on the liver surface) and WB-DWI/MRI estimated a PCI of 19 (partially shown with the arrowheads in B as diffuse confluent metastases at the liver surface). The high PCI of 19 was confirmed during explorative laparotomy. The patient was deemed inoperable and received palliative chemotherapy

Fig. 3

A 51 year old male patient with cancer at the rectosigmoid junction and clinical suspicion of peritoneal metastases underwent a WB/DWI-MRI (a/b) in addition to a CT scan (c) to evaluate operability. WB/DWI-MRI showed extensive peritoneal cancer spread, partially shown by the white arrowheads in the zoomed-in image (b) of the coronal b1000 DWI (a), with an estimated PCI of 26. The tumour deposit at the splenic flexure of the colon (arrow in b) was also recognized on CT, shown by the white arrow on the coronal CT image (c), but the other lesions were not seen on CT. Disease load was clearly underestimated with an estimated PCI of 8. At explorative laparotomy extensive peritoneal disease was confirmed (PCI 33) and the patient received optimal palliative care

Fig. 4

A 49 year old male patient with a history of a sigmoid resection because of adenocarcinoma was diagnosed with tumour recurrence. He had a WB-DWI/MRI for operability assessment. Apart from limited peritoneal disease (not shown on the images) a liver metastasis was found in liver segment 4B, shown with the arrows on the coronal b1000 DWI (a) and the coronal T2-weighted image (b). This liver metastasis was not recognised on the axial (c) and coronal (d) CT images. The patient could undergo a curative debulking with RFA of the liver metastasis