Radiologic Assessment

Good-quality cross-sectional imaging of the liver — in some cases, the abdomen and chest — is required to evaluate the number of tumors, their size and location, relative to the structures within the liver. The most common cross-sectional imaging used is CT scan. The quality is excellent and it is the least expensive option. It must be a contrast-enhanced CT scan, high-quality cross-sectional imaging. CT scanning with contrast is one of the mainstays of imaging before considering surgical therapy on the liver.

One could use contrast-enhanced MRI in some cases — it depends on the institution. I don’t think MRI adds that much more, but in some centers it’s helpful. In some cases, like with chemoembolization or with PVE, MRI is taken because the contrast material that’s injected to perform these intra-arterial or intravascular procedures can result in a poor-quality CT. Therefore, occasionally we’ll opt for MRI.

In addition, we generally recommend a PET scan for patients undergoing surgical therapy for the liver, because PET is more sensitive at picking up metastatic disease both within and outside the liver, in peritoneal implants, lymph nodes and other metastases in the body. In contrast to so-called structural scanning, it is functional imaging. In one out of four patients, findings from a PET scan can prevent the use of unnecessary surgery or unneeded therapy. We generally recommend high-quality cross-sectional imaging, typically a PET/CT scan.

The PET/CT is a combination of a PET scan and a CT scan, but a different type of CT scan that is often done without contrast. Indeed, the CT part of a PET/CT does improve the accuracy of the PET scan, because it’s a combined scan that shows more anatomy or more structure, so it improves the accuracy of the PET scan. However, it does not produce the same results as the contrast CT scan, unless you specify that you want to do a contrast CT as part of the PET/CT.

Volumetrics is another technique in which the radiologist can almost perform virtual surgery, prior to the real surgery, to map out the remnant volume, calculating the volume. That’s a variation of the CT scan that is rarely needed, except for selected patients with borderline resectable status.

High-quality intraoperative ultrasound is used by the surgeon to guide the surgery.

Clinical implications of complete radiologic response

Commentator: NICHOLAS PETRELLI, MD

For patients with liver metastases who are treated with chemotherapy and targeted agents and, consequently, have a complete imaging response on CT, the data show that 80 percent still have disease in the liver even though you can’t see it on imaging.

It is rare to achieve a complete pathologic response with the agents that we have today. In my mind, even if the imaging shows a complete response, you still have to remove those areas of the liver. Otherwise, you’re going to leave tumor behind in those patients. A couple of months down the line, the tumor’s going to show up and people are going to say that’s a recurrence, but in actuality it was residual cancer.

Scanning techniques to evaluate patients with liver metastases

Commentator: JOHN PRIMROSE, MD

Comment:

For all patients, we use a multislice CT of the chest, abdomen and pelvis, and MRI scans are performed on all patients who will undergo liver resection. In our practice, we do not perform PET scanning on everyone — we use PET scans with patients who are likely, on the basis of the presentation, to have extrahepatic disease. So that is the value of PET imaging. It’s unusual for PET to provide much information about the liver, unless there is some equivocation from the liver-dedicated CT as to whether a lesion is malignant or not. In that situation, it sometimes helps, but normally a PET scan is used to evaluate the patient for potential extrahepatic disease. CT and MRI provide more information on the liver than the PET/CT. The other problem we have with PET/CT is that there are false-positives and false-negatives. In particular, we’ve found patients who have inflammatory fossae in the pelvis that are called malignant by the report and turn out to be benign. So if we find something unexpected on a PET scan, we always follow up with some further imaging and probably a biopsy, to ensure that it wasn’t a false-positive. Otherwise, you may be denying a patient an operation that could be beneficial on the basis of our investigation. I think PET/CT isn’t used as much in the UK liver centers, as in the United States, where I think it’s more commonly used.

Obtaining a good radiologic margin

Commentator: JOHN PRIMROSE,MD

Comment:

The definition of a good radiologic margin is controversial. Given the choice, we would all prefer to have a liver margin clear of any metastatic deposit on our excision specimen — that would be the ideal, but often that’s not possible. Current thinking is that any margin is probably acceptable — if you can’t obtain anything better — provided you remove the tumor. Even if it’s a millimeter or less, it’s probably a reasonable thing to perform, because not all patients will go on to develop local recurrence. Many of us use an ultrasound dissector during liver surgery, which probably produces a margin of a millimeter or so, in addition to what the pathologist receives. So when we talk about margins, we’re not always talking about the same thing.

In terms of anatomy, we obviously are concerned about the inflow and the venous drainage, so the bile duct usually doesn’t matter so much, because you can replace it. Sometimes you can resect and reconstruct the portal vein. The hepatic artery is much more difficult to work around. Margins on the major vascular structures are what cause us the most concern. Most of us would say that any margin is better than nothing. Even if you can’t obtain a centimeter, it’s probably still worth performing.

Image quality and assessment of disease recurrence

Commentator: JOHN PRIMROSE, MD

Comment:

It is difficult to predict prognosis according to the time from adjuvant chemotherapy to recurrence. I think the difficultly in a lot of situations lies in the quality of the imaging that was collected up front. If you obtain high-quality imaging for these patients, in all probability you’ll detect that the disease has always been synchronous in presentation. A better scan will reveal more than a poor scan. I think up front you might say, “A patient who gets disease that quickly after adjuvant chemotherapy is likely to have worse disease.” That may be true, but I don’t think we can prove that, and it certainly does not affect my decision-making process in any way about offering them surgery.

Typical serum and imaging surveillance

Commentator: STEVEN CURLEY, MD

Comment:

In most patients with Stage II colon cancer, many physicians only follow serum CEA after two to three years. Many feel that the cure rate for patients with Stage II tumors who receive proper surgical therapy and adjuvant therapy is high, so only standard chest radiographs are used. Some physicians would simply evaluate the CEA level on a semiannual or annual basis. I would have to say that this is an area of continued controversy and that there’s no ready agreement. We tend to be a little bit more aggressive in our follow-up at MD Anderson. We will perform CT scans at least annually until a patient reaches five years out, but that’s not certainly agreed on by all practitioners.

Surveillance after primary therapy for patients with Stage II or III colon cancer

Commentator: AXEL GROTHEY, MD

Comment:

The way we generally monitor patients with Stage II disease is we begin with a baseline CT scan and CEA after surgery, and then repeat both every six months. With Stage III disease, we obtain the baseline studies and then repeat both every three to four months for the first two years, and then every six months thereafter.

Reliability of PET scans in patients receiving chemotherapy

Commentator: AXEL GROTHEY, MD

Comment:

When PET scans are performed on patients receiving chemotherapy, we commonly see false-negatives. The reason for that is a PET scan utilizes the functional capability of tumor cells to metabolize glucose. In response to chemotherapy, cells might shut down their metabolism and enter a dormant stage — they are not PET-positive, but they are still alive.

Emerging data in various tumors, and potentially in colon cancer, show that if you administer treatment and see a response on the PET scan, you can repeat the scan two weeks later to identify early responders. This indicates better overall prognosis, which could make sense, because early responsiveness to therapy might indicate this tumor is extremely chemotherapy sensitive.

However, after say three months of therapy, most tumors have a PET response. You may think these tumors are dead, but they’re not dead, so I normally do not take into account PET response after a longer period of time, and particularly not in a patient who has resectable disease.

Radiologic Disease Assessment