Abstract
Advanced primary hepatocellular carcinoma (a-PHC) has reached almost epidemic proportions in the Third World where, with a median survival of about 5 months, its management is draining health care resources, without, arguably, any economic returns. Searching for a pragmatic approach to the management of patients with clinical features suggestive of advanced a-PHC, we have performed a decision analysis and plotted a decision tree structure based on literature review with the focus on serum alpha-fetoprotein (AFP), and the expected outcome from liver biopsy, and chemotherapy. We defined serum AFP of <40 ng/ml as negative, and, 40 ng/ml as positive. The results were as follows. The probability of such a patient having a-PHC would be 0.19 or 19% if his serum AFP is negative; conversely, a positive serum AFP confers a 0.84 or 84% probability of him having PHC. Put differently, the likelihood ratios of PHC are 0.24 and 5.33 for serum AFP negative and positive cases, respectively. Blind percutaneous liver biopsy in patients with a-PHC has a low accuracy, or it involves a high risk, or both. The decision tree structure showed that the average expected utility from chemotherapy (3.12) and supportive care only (3.2) would be similar. We conclude that, given the current impasse in the management of such patients, the following pragmatic approach is justifiable. Serum AFP should be done, but neither liver biopsy nor chemotherapy is worthwhile in patients with a-PHC. We suggest that, in Third World countries, greater emphasis should be placed on the prevention of PHC rather than on futile aggressive therapy.