We thank Dr. Tez and colleagues for their interest in our study1 and for their remarks. Dr Tez suggests that slight variations in population characteristics can change the predictive value of the scoring system, and the use of the neurofuzzy model can improve the predictive values. In our previous report, we demonstrated that the Cancer of the Liver Italian Program (CLIP) was the best stratification system for patients with hepatocellular carcinoma (HCC) undergoing transarterial chemoembolization.1 Of course, some variations in population characteristics may change the predictive values of the CLIP system. However, significant factors determined by the use of multivariate analysis in our study were very similar to those factors constituting the CLIP score.1
In my opinion, the neurofuzzy model seems to be an open system that can encompass any variable as a prognostic factor of survival.2, 3 It may provide a better risk estimation for survival, but I would like to comment on some potential limitations of the system as follows. First, several dozens of variables can be incorporated into the scoring system.3 For practitioners, it would not be practical to consider all of these variables together for a prediction of prognosis for each patient. The simpler system is the better one to use as a scoring system. Second, as the number of variables increases, the penalty for the Akaike Information Criterion scores becomes larger. Accuracy and simplicity should go together with regard tothe scoring system. Third, the prognostic factors would be different for each developing institution of the neurofuzzy model and for different populations. International agreement for the liver staging system is essential as a practical point. Finally, despite the complexity of the system, little information can be derived for the choice of treatment option.
I believe that the neurofuzzy model can be an effective tool in the development of any scoring system. However, while considering the above-mentioned potential limitations, I do not believe that the model can replace the current simple and practical liver staging systems that are used worldwide.