Reply: Selection Criteria for Liver Transplantation in Patients with Hepatocellular Carcinoma: Beyond Tumor Size and Number?

We appreciate the interest of Dr. F. Yao, whose editorial assessing the problems of liver transplantation (LT) for hepatocellular carcinoma included in particular a discussion of our article published in the August issue of Liver Transplantation.1, 2 Indeed, transarterial catheter embolization (TACE) before LT is not new. It has been performed by several transplantation centers.3-6 The new aspect in our approach is that patients exceeding the Milan criteria7 were downstaged,8 and, in contrast to any other study, the patients listed for LT were continuously treated by TACE during the waiting period according to a strict protocol. Risk of tumor recurrence was low in the patients without measurable tumor progress compared with patients with minimal tumor progress during the waiting time, even if this “progress” remained within the official definition of stable disease.8 These results did not depend on whether the patients were downstaged to within the Milan criteria. The biological selection pressure by TACE maintained during waiting time led to these results. Biological tumor behavior rather than tumor size controls the risk of disease recurrence. After this pretreatment, the generally accepted prognostic criteria appear to be no longer relevant. Grading and microvascular invasion correlated with tumor diameter in many studies7, 9-11 including the study of Mazzaferro et al.,7 which assessed small and oligocentric tumors meeting the Milan criteria; poorly differentiated tumors and microvascular invasion were not addressed in their study. In the 16 patients in our study who met the Milan criteria, 5 tumors were poorly differentiated, and 3 had microvascular invasion. Are these discrepancies caused by incidence, by different pathomorphological assessment and diligence, or by patient selection due to TACE? The studies performed at the University of California at San Francisco11, 12 showed the strong prognostic influence of tumor burden on the bases of size and number, and the influence of downstaging to a stage meeting the Milan criteria. However, the pretreatment in this study was different from our approach. We therefore assume that the negative influence of the number of tumor nodules on survival in our study is not caused by tumor burden but by the biological aggressiveness of these tumors, leading to multiple intrahepatic tumors and to systemic disease responsible for recurrence after LT. In his review, Yao raised a critical problem: the reliability of the assessment of stable disease and minimal progress. The decision to characterize a situation as “minimal progress” if this “progress” is defined by official criteria as stable disease8 may be arbitrary. The limitation of imaging techniques contributes to the difficulties with this decision. If the criterion “minimal progress during waiting time” is used in clinical practice, then a patient may be excluded from LT who should have undergone transplantation, or vice versa. Is it justifiable to exclude a patient whose tumors increase in size from 2 to 3 cm while the patient remains on the waiting list, despite pretreatment with TACE? This issue could perhaps be answered if patients with tumors meeting the Milan criteria are included in a simple nonrandomized study. What we need is a great number of patients within the Milan criteria. These patients have to be pretreated according to our TACE protocol. Any patient—regardless of whether his or her tumors are stable (regressive) or progressive—remains on the waiting list as long as the tumor does not exceed the Milan criteria. Stable disease or progress have to be prospectively assessed, and recurrences after transplantation need to be recorded. We would expect tumor recurrence to be different in these 2 groups. Gerd Otto*, * Department of Transplantation and Hepatobiliary Surgery, University of Mainz, Mainz, Germany.