Add liver MRI to CT before resecting pancreatic cancer

Design

Systematic review and bivariate meta-analysis of direct comparative studies

Population

40 studies of patients with suspected pancreatic ductal adenocarcinoma

Primary outcome

Sensitivity and specificity for detection and staging

Effect

Liver metastases: MRI sensitivity 0.89 vs CT 0.60; resectability specificity CT 0.88 vs MRI 0.63

This PROSPERO-registered meta-analysis included 40 studies in which patients with suspected pancreatic ductal adenocarcinoma had at least two imaging tests, with pathology as reference where available. Direct comparison within the same patients reduces the bias that comes from comparing different populations.

CT, MRI and endoscopic ultrasound did not differ in detecting the primary tumour. For liver metastases MRI was clearly better: sensitivity 0.89 (95% CI 0.67 to 0.97) and specificity 0.91 (0.76 to 0.97), against 0.60 (0.44 to 0.74) and 0.84 (0.70 to 0.93) for CT. CT was better for vascular invasion and more specific for resectability (0.88, 0.76 to 0.94, vs 0.63, 0.48 to 0.76). Endoscopic ultrasound was more sensitive than CT or FDG-PET/CT for nodes.

The two tests answer different questions. CT remains the study for local vascular staging and distant disease, and MRI of the liver finds the small metastases CT misses, which would otherwise be discovered at laparotomy or within months of a futile resection. Where MRI access is limited, as in many Indian centres, it is most worth prioritising for patients whom CT has called resectable.

• Keep pancreatic protocol CT as the primary staging test for vascular involvement and resectability.
• Consider liver MRI before potentially curative pancreatic resection.
• Prioritise MRI for patients whom CT has staged as resectable or borderline resectable.
• Use endoscopic ultrasound where nodal assessment or tissue diagnosis is needed.
• Mention in the CT report when liver lesions are too small to characterise.

Why it matters

A resection performed in the presence of missed liver metastases exposes the patient to major surgery without the chance of cure.

Don't overread it

MRI did not outperform CT overall; its advantage is specific to liver metastases.

The statistics, in plain English

A CT sensitivity of 0.60 means CT missed about four in ten patients' liver metastases, against about one in ten for MRI. The intervals are wide because fewer studies reported this outcome, but they barely overlap. A resectability specificity of 0.63 for MRI means it more often wrongly labels resectable tumours as unresectable, which is why CT stays first for that question.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.