It makes sense to screen for liver cancer just as we do for colon, cervical, breast, and prostate cancer. However, the difference is that there is, as yet, no cost-effective way of screening for liver cancer. Blood levels of alpha-fetoprotein are normal in up to 50% of patients with small liver cancer. Ultrasound scanning, which is non-invasive and very safe, is, as mentioned before, operator-dependent. Therefore, the effectiveness of a screening ultrasound that is done at a small facility can be very suspect.

Even more disappointing is the fact that no study outside of Asia has shown, on a large scale, that early detection of liver cancer saved lives. Why is that? It is because, as already noted, the treatment for liver cancer, except for liver transplantation, is not very effective. Also, keep in mind that patients found with small tumors on screening live longer than patients with larger tumors only because of what is called a "lead time bias." In other words, they seem to liver longer (the bias) only because the cancer was discovered earlier (the lead time), not because of any treatment given.

Nevertheless, strong arguments can be made for routine screening. For example, the discovery of an liver cancer in the early stages allows for the most options for treatment, including liver resection and liver transplantation. Therefore, all patients with cirrhosis, particularly cirrhosis caused by chronic hepatitis B or C, hemochromatosis, and alcohol, should be screened at six- to 12-month intervals with a blood alpha-fetoprotein and an imaging study. I favor alternating between an ultrasound and CT scan (or MRI). Patients with chronically (long duration) elevated alpha-fetoprotein levels warrant more frequent imaging since these patients are at even higher risk of developing liver cancer.