The letter by Caturelli and de Sio allows us to clarify several controversies regarding the diagnostic usefulness of fine-needle biopsy (FNB) of nodules detected during ultrasound (US) screening in cirrhosis and address some of their concerns and misunderstandings. (1) The aim of our study was designed to prospectively validate the noninvasive diagnosis of hepatocellular carcinoma (HCC) in nodules smaller than 20 mm as proposed by the American Association for the Study of Liver Diseases (AASLD), and the methodology is described in ample detail in the manuscript.1 We included newly detected, solitary nodules between 5-20 mm in patients with cirrhosis and describe the diagnostic work-up both in text and figure including the data obtained by imaging techniques (contrast-enhanced US [CEUS], dynamic magnetic resonance imaging [MRI], and FNB) both at baseline and during follow-up. Biopsy was used as the gold standard for nodules in a cirrhotic liver when this could be done safely. We based the final diagnosis on pathology in all cases, except in five individuals in whom biopsy was not feasible (three cases in whom the nodules were no longer visible at the time of biopsy) or not justified (two cases with unequivocal diagnosis of hemangioma at MRI). There was a single patient in whom the nodule growth was beyond 2 cm during the work-up and who displayed intense arterial uptake followed by venous washout on MRI despite two consecutive negative biopsies. CEUS was not possible because of lung interposition but computed tomography scan reproduced the MRI findings. From an ethical perspective, it was not justified to obtain a third biopsy and delay treatment decision in a patient in whom the nodule evolved into a lesion fitting the HCC profile according to available guidelines.2, 3 Our patient was treated by ablation and is now on the waiting list for transplantation. Clinical research should balance the clinical practice needs and patient protection versus a study design that could be readily implemented in experimental models. Surely Caturelli and de Sio agree with these arguments. (2) The second misunderstanding by Caturelli and de Sio refers to the applicability of FNB. As clearly stated in the manuscript, we included patients who had nodules amenable to biopsy both in terms of location and lack of contraindications (such as clotting disorders or ascites, circumstances that are common in patients with advanced cirrhosis). These restrictive inclusion criteria explain the good applicability of biopsy in our series, and the sole patients without biopsy are those diagnosed with hemangioma or in whom the nodules vanished (this has already been discussed in point 1). Being aware of the problems of biopsy in patients with cirrhosis, it is indeed surprising that Caturelli and de Sio had 100% applicability in their studies, where no restriction was applied.4 Experienced interventional radiologists acknowledge that there are patients in whom biopsy can not be obtained, and this is one of the concerns of the data they have published, in addition to the high diagnostic accuracy of biopsy in their studies (see point 3). (3) Caturelli and de Sio claim that our diagnostic accuracy with biopsy is too low and that their experience is far more positive. Their technique of obtaining cytologic and histologic preparations, even if using fine-needle aspiration, is conventional in all units and should not be the reason for the differences. In fact, in our protocol we planned both fine-needle aspiration and FNB by two separate punctures that in half of the nodules were not feasible because of their location. In the Caturelli study, with nodules smaller than 2 cm in size, a sensitivity of more than 68% is reported in nodules <10 mm and sensitivity of more than 91% in those exceeding 10 mm.4 This agrees with the rate of false negatives in the biopsy. This well-known issue in HCC diagnosis by biopsy can only be properly addressed with the use of an adequate gold standard. The 12% failures (false negatives) in the study by Caturelli et al. have been registered as HCC because of the appearance of hypervascularization in any imaging technique in nodules smaller than 2 cm. The European Association for the Study of the Liver guidelines that the authors claimed to have used did not allow them to establish HCC diagnosis through imaging techniques (biopsy was the sole diagnostic criteria at that time) and the AASLD guidelines and the results of our data indicate that coincidental findings are required to prevent false positive HCC diagnosis. Of course, we have to acknowledge that even with these contentions, the Italian multicenter survey has a number of false negatives, but their sensitivity results and applicability would still be unique and raise concerns when compared with data from the rest of the groups in Europe, Asia, or the United States with expertise in the field. An important note is the low prevalence of non-HCC diagnosis in the population studied by Caturelli et al., because this may also suggest a selection bias in avoiding nodules highly unlikely to be HCC and/or a potential overdiagnosis of HCC for any reason. (4) Finally, the main objective of our study was not to eliminate the value of biopsy, but to validate the accuracy of the noninvasive criteria. As previously mentioned, there are several reasons why biopsy may not be performed, and in these cases the AASLD definitions are of high clinical value. Similarly, it could be argued that if the noninvasive criteria are met, biopsy and the associated risks of bleeding and seeding can be avoided. On the contrary, if the criteria are not met the physicians and patients have to balance the risk of biopsy with the risk of engaging a treatment (transplantation) because of an unconfirmed malignancy. Caturelli and de Sio claim that there are no data about complications related to the biopsy in our study, an omission that is also observed in their study. Because description of complications was not the aim of the study, we did not focus on it and we are happy to state that we had no bleeding complications in our study. Detection of seeding would require longer follow-up. At the time of publication, there was no known seeding; we now have one of these events in a patient whose HCC was treated by ethanol injection. Regarding the comments by Song, we would like to stress that the gold standard used in our study was the result of the FNB (and not the histological analysis of the resected specimen). We agree with Song that confident diagnosis of HCC through a FNB in these small nodules is difficult, due to sampling error and the difficult distinction between high-grade dysplastic nodules (HGDNs) and early HCC. These justify the false negative results obtained by the study. In addition, although some of the dysplastic nodules could have been interpreted as HGDNs (or indeed early HCC) by the Japanese pathologist, all these benign nodules were closely followed-up during a median time of more than 24 months and we were able to confidently discard a diagnosis of HCC. Finally, we included patients with solitary, newly detected nodules detected by US screening that were well-defined and identified in two US projections. By using these requirements, we avoided the inclusion of patients with coarse, ill-defined, previously present nodules that correspond to regenerative nodules within a cirrhotic liver. If we had included patients with several nodules, the likelihood that one of them would correspond to HCC would have been higher. If we had to study them all at the same time, we would have had to biopsy all of them, which is a proposal that would not be acceptable, as Song will surely agree. In summary, we hope that these comments have helped to clarify some of the controversies raised by Caturelli, de Sio, and Song, while also exposing the limitations and benefits that FNB offers in the diagnosis of early HCC. Current research efforts in genomics and proteomics may overcome the need of both pathology interpretation and imaging characterization, but in current clinical practice both tools have to be used in a complementary way. Alejandro Forner*, Jordi Bruix*, * Barcelona-Clinic Liver Cancer (BCLC) Group, Liver Unit, Hospital Clinic, Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas (CIBEREHD), Institut d'Investigacions Biomèdiques August Pi i Sunyer, University of Barcelona, Barcelona, Spain.