'Early detection of hepatic metastasis in follow-up high-risk colorectal carcinoma*

Each year, in the Netherlands 15,500 people develop colorectal carcinoma and
about 5,100 patients die of this condition, particularly due to the development
of metastases.
Approximately 50-60% of patients will develop liver metastases. The risk of
developing liver metastases is highest in colorectal carcinoma patients with
positive locoregional lymphnodes, the so-called high-risk colorectal carcinoma
patients (pN1 / N2). About 40-50% of these patients will develop metastases
within 3 years.
When patients develop liver only metastases, local treatment with curative
intent is preferred. After resection of the liver metastases, the 5-year
survival is 40-50%. However, we assume the earlier liver metastases are
diagnosed, the higher the chance a curative resection can be performed.
To detect liver metastases as early as possible, all patients with colorectal
carcinoma are followed for at least 5 years. Six, 12, 24 and 36 months after
surgery an ultrasound of the patients liver is performed. Furthermore every 6
months the tumor marker CEA in blood is determined.
The sensitivity of ultrasound for the detection of liver metastases is 57%².
The sensitivity of MRI for the detection of liver metastases is 88%³. However a
MRI of the liver is costly and time consuming compared to ultrasound. The
estimation is that a shortened MR protocol of the liver (saves money and time,
compared to an extensive protocol) increases the sensitivity for the detection
of colorectal liver metastases, compared to ultrasound. This means that liver
metastases are detected earlier. We assume the earlier liver metastases are
diagnosed, the chances of curation will increase. The hypothesis is that
earlier detection of a first liver metastasis (LM 1) can reduce the risk to and
extend time to develop/diagnose (possibly) recurrent liver metastasis (LM 2).

1. The primary objective of this study is to investigate if follow-up of
asymptomatic patients with high-risk colorectal carcinoma with reduced protocol
MRI liver instead of US will affect the time to diagnosis recurrent liver
metastasis (LM-2) in months starting from postoperative control
(=randomization); TRLM.

Secondary objectives are to investigate if follow-up of asymptomatic patients
with high-risk colorectal carcinoma with reduced protocol MRI liver instead of
US will affect:
2. Time to diagnosis of first liver metastasis (LM-1) in months starting from
postoperative control (=randomization); TFLM.
3. Time between first postoperative control and diagnosis of first liver
metastasis (LM-1); liver metastasis free survival 1 (LMFS-1).
4. Time between diagnosis LM-1 and diagnosis LM-2 in months; liver metastasis
free survival 2 (LMFS-2).

5. The proportion of patients who are potentially eligible for curative
therapy when detecting LM-1.
6. 5 year survival.
7. Anxiety and quality of life.

8. To compare sensitivity and specificity for detection of liver metastasis
between US and reduced MRI liver protocol.

This study is designed as randomized, single blinded, parallel group controlled
trial.

Patients in the control group will undergo US at regular intervals (6,12,24 and
36 months after surgery) as part of a common care. Besides these US*s the
intervention group will also undergo a shortened MRI liver protocol at these
intervals. This is necessary to obtain a comparison of sensitivity and
specificity between US and the shortened MRI protocol.

In the intervention group besides regular US an MRI of the liver will be
performed. The potential risk outweighs the likely benefit.