[Role of imaging in abdominopelvic follow-up after resection of colorectal cancer].

The imaging techniques used to follow-up patients submitted to surgery for colorectal cancers are presented as are their advantages and disadvantages. The sensitivity of ultrasonography (US) for postoperative surveillance of the liver varies considerably according to whether the absence or presence of metastases is sought (83%) or their exact number (53-82%). US is recommended by French clinicians as the standard follow-up examination whereas, American clinicians tend to discuss CT-Scan or MRI for the same purpose. Some authors advocate Doppler ultrasonography for the detection of liver metastases. Spiral CT-Scan offers better results than conventional CT and is now considered second in rank to US for follow-up studies. Its sensitivity exceeds 90% for metastases greater than 1 cm. IRM and arterial CT-scan are not applicable for follow-up but are sometimes indicated for difficult cases before hepatectomy. US is also an appropriate technique for postoperative exploration of the abdominal cavity. When the CEA level and the US examination are normal, this technique is sufficient. However, in dubious clinical or biological cases, US should be associated with a spiral CT-Scan. Most clinicians recommend a reference CT-Scan, two to four months after the resection and every 6 months for 2 years and then annually. The same follow-up strategy is advocated after abdominoperineal resection and for surveillance of the pelvis. The sensitivity of CT-Scan is between 69 and 88% but it is unable to differentiate recurrence from fibrosis. IRM, with a sensitivity of 90%, is the best method for overcoming with problems related to the differential diagnosis and particularly novel fast acquisition imaging and contrast agents. Ultrasonoendoscopy, repeated every 4 to 6 months after anterior resection, is also a useful examination for surveillance of the pelvis. PET-Scan has produced encouraging results for the detection of metastases and for the differential diagnosis between local recurrence and fibrosis. When available, this technique could become the "gold standard" for the surveillance of patients following surgery for colorectal cancers.