ORIGINAL ARTICLE. Impact of Histological Grade of Hepatocellular Carcinoma on the Outcome of Liver Transplantation

Transcription

1 ORIGINAL ARTICLE Impact of Histological Grade of Hepatocellular Carcinoma on the Outcome of Liver Transplantation Sumihito Tamura, MD; Tomoaki Kato, MD; Mariana Berho, MD; Evangelos P. Misiakos, MD; Christopher O Brien, MD; K. Rajender Reddy, MD; Jose R. Nery, MD; George W. Burke, MD; Eugene R. Schiff, MD; Joshua Miller, MD; Andreas G. Tzakis, MD Hypothesis: Histological grade of hepatocellular carcinoma (HCC) is an important prognostic factor affecting patient survival after orthotopic liver transplantation (OLT). Design: Retrospective analysis. Setting: University-based teaching hospital. Patients: Of 952 OLTs performed between June 1991 and January 1999, 56 OLT recipients had histologically proven HCC in the explant liver. Of those, 53 patients with complete clinicopathologic data were analyzed. A single pathologist blinded to the outcome of each patient reviewed all histological specimens. Results: Median follow-up was 79 days. Overall survival for patients with tumors sized 5 cm or less at 1, 3, and 5 years was 87%, 78%, and 71%, respectively (Kaplan-Meier). Univariate analysis revealed the size, number, and distribution of tumors; the presence of microscopic vascular invasion and lymph node metastasis; histological differentiation; and ptnm stage to be statistically significant factors affecting survival. Multivariate analysis revealed histological differentiation and ptnm stage to be the independent and statistically significant factors affecting survival (P=.2 and.3, respectively). When ptnm stage was excluded from multivariate analysis, histological differentiation and size remained the significant independent factors (P=.2 and.3, respectively). Threeyear survival for patients with small ( 5 cm) tumor with well- to moderately differentiated and poorly differentiated HCC was 82% and 67%, respectively. Three-year survival for patients with large ( 5 cm) tumor with well- to moderately differentiated and poorly differentiated HCC was 62.5% and %, respectively. Conclusions: In our retrospective experience, histological differentiation had a statistically significant effect on the prognosis of HCC after OLT. However, before altering the current OLT selection criteria for patients with HCC, prospective studies are required to confirm the impact of histological grade on clinical outcome. Arch Surg. 21;136:25-3 From the Division of Transplantation, Department of Surgery (Drs Tamura, Kato, Misiakos, Nery, Burke, Miller, and Tzakis), the Division of Immunopathology, Department of Pathology (Dr Berho), and the Division of Hepatology, Department of Medicine (Drs O Brien, Reddy, and Schiff), University of Miami School of Medicine, Miami, Fla. HEPATOCELLULAR carcinoma (HCC) is a major cause of death worldwide, with a high incidence in Asia and South Africa. 1 More than 8% of patients with HCC have liver cirrhosis, particularly related to chronic viral hepatitis types B and C. 2 Although surgical resection of the tumor is considered the best option, 3 it is possible in only a minority of patients owing to impaired functional reserve of the liver and multicentricity. To overcome the limited application of resection, other treatment modalities have been investigated. 4,5 Use of orthotopic liver transplantation (OLT) for the treatment of HCC has been controversial. Although the results to date are comparable to those for conventional partial hepatectomy, 6,7 the application of transplantation for treating cancer remains a difficult decision given the shortage of donor organs. Earlier studies 6,8 observed that patients with HCC did poorly compared with patients with benign disease. A recent experience, 9 however, reported that patients with HCC, who were carefully selected on the basis of the size and number of lesions, did as well after OLT as patients without cancer. See Invited Critique and Invited Response at end of article In addition, data from a large international registry 1 suggested that histological differentiation of the tumor was an important prognostic factor in the outcome of OLT for HCC. We describe our experience evaluating multiple factors, including clinicopathologic features, in the outcome of HCC after OLT. 25

2 PATIENTS AND METHODS Between June 1991 and January 1999, 952 liver transplantations were performed at Jackson Memorial Hospital/University of Miami Medical Center, Miami, Fla. A retrospective study revealed 56 OLT recipients with histologically proven HCC in the explant liver. Patients with so-called incidental tumor, ie, tumor not identified during pretransplant workup but found in explant liver, were also included in the study. Histological slides from 53 of the 56 patients were available for analysis, and these patients formed the study population. Data on demographics; pretransplant -fetoprotein level; Child-Pugh class; presence of viral hepatitis types B and C; chemotherapy before and after OLT; size, number, and distribution of tumors; ptnm stage according to Union Internationale Contre le Cancer (UICC) criteria 11 ; histological differentiation of the tumor; presence of microscopic vascular invasion and lymph node metastasis; and donor bone marrow infusions were collected and analyzed. Histological data on the tumor lesions of all patients were retrieved and reviewed by a single expert pathologist (M.B.) masked to the data on the outcome of patients after undergoing OLT. Histological differentiation was defined by the criteria of Kondo 12 and Sugihara et al, 13 which classify HCC into 3 categories: well, moderately, and poorly differentiated. These classifications have overlapping and similar definitions to the Edmondson-Steiner classification. 14 Overall survival and disease-free survival rates were calculated using the Kaplan-Meier method, 15 and statistical significance was defined using the logrank test. 16 Factors related to survival were analyzed using the Cox proportional hazards regression model 17 and SAS statistical software (SAS Institute Inc, Cary, NC). Statistical significance was defined as P.5, corresponding to the test of null hypothesis that all coefficients associated with the risk factor in the regression are zero. Comparison of continuous ordinal data between 2 groups was performed using the Mann-Whitney U test. RESULTS DEMOGRAPHIC DATA Of 53 patients in the study, 36 (68%) were men and 17 (32%) were women. Mean age was 57 years (range, 2-77 years). Most patients (n=44 [83%]) had viral hepatitis, 37 (7%) of whom were hepatitis C virus antibody positive. Twenty patients (38%) had Child-Pugh class B and 2 (38%) had Child-Pugh class C cirrhosis. Mean tumor size was 4.4 cm (range,.5-18 cm). After OLT, HCC (incidental tumor) was found in the explant livers of 13 patients (25%). Mean tumor size in incidental cases was 2.5 cm (range,.5-6. cm). Twenty-four patients (45%) received donor bone marrow infusions for induction of immunologic tolerance after OLT. Other demographic features are summarized in Table 1. Table 1. Overall Survival According to Clinicopathologic Features* Overall Survival, % Factor Patients, No. 1y 3y Age, y Sex Male Female Viral hepatitis Yes No Incidental Yes No Fetoprotein, µg/l Child-Pugh class A and B C Size, cm Multiple tumors Yes No Bilobar tumors Yes No Vascular invasion Yes No Lymph node positive Yes 2 No Histological differentiation Well and moderate Poor UICC ptnm Stage II Stage III Bone marrow infusion Yes No Chemotherapy before OLT Yes No Chemotherapy after OLT Yes No SURVIVAL AND RECURRENCE Median follow-up was 79 days and mean follow-up was 887 days, with the longest survival exceeding 8 years (2976 days to date). Of 53 patients, 17 (32%) died and 36 (68%) are currently alive. Four patients died of systemic infection, 1 of chronic rejection, and 1 of an acute cardiac event. None died of recurrent hepatitis. Overall survival at 1, 3, P *UICC indicates Union Internationale Contre le Cancer; OLT, orthotopic liver transplantation. Kaplan-Meier method. By log-rank test for overall patient survival

3 Cumulative Survival and Recurrence % 79% 1 y 65% 6% 6% 3 y 5 y Overall Survival Disease-Free Survival 61% Follow-up, d Table 2. Cox Regression Model for Histological Differentiation and ptnm Stage and for Histological Differentiation and Size* Factor Relative Risk (95% CI) P Analysis With ptnm Stage Histological differentiation, well 3.49 ( ).2 and moderate vs poor ptnm stage, II vs III 5.65 ( ).3 Analysis Without ptnm Stage Histological differentiation, well 5.16 ( ).2 and moderate vs poor Size, 5vs 5cm 2.96 ( ).3 Figure 1. Overall survival and recurrence in 53 patients. Median follow-up was 79 days and mean follow-up was 887 days, with the longest survival exceeding 8 years (2976 days). No recurrence was noted more than 3 years after orthotopic liver transplantation. *CI indicates confidence interval. Relative risk of disease recurrence estimated in multivariate analysis using forward stepwise procedure in Cox regression model. and 5 years was 79%, 65%, and 61%, respectively, whereas survival for patients with tumor size of 5 cm or less was 87%, 77%, and 71%, respectively. One-, 3-, and 5-year disease-free survival was 77%, 6%, and 6%, respectively (Figure 1). As of January 1999, there was no evidence of recurrence of HCC in 15 patients who had survived longer than 3 years. Of 14 patients (26%) in whom HCC recurred, 11 died of metastatic disease. One patient with recurrence died of systemic infection. Two patients who experienced recurrence were alive and well as of January Liver (n=4 [29%]), lung (n=3 [21%]), and bone (n=2 [14%]) were the common sites of recurrent or metastatic disease. At the end of the study period, 34 patients (64%) were alive and free of disease. Proportion Survival % at 3 y Well- to Moderately Differentiated HCC (n = 8) Poorly Differentiated HCC (n = 6) Follow-up, d UNIVARIATE ANALYSIS FOR SURVIVAL Univariate analysis regarding survival was performed with the following factors: age, sex, presence of viral hepatitis, incidental tumor, pre-olt -fetoprotein level, Child-Pugh class, size of the tumor, presence of multiple tumors, presence of bilobar tumor, microscopic vascular invasion, microscopic lymph node metastasis, histological differentiation, UICC ptnm stage based on microscopic findings, donor bone marrow infusion, and chemotherapy(transarterial chemoembolization and intravenous chemotherapy). Eight factors incidental tumor, size, multiple tumor, bilobar tumor, vascular invasion, lymph node metastasis, histological differentiation, and ptnm stage significantly affected survival (P.5, log-rank test). Bone marrow infusion after OLT did not affect survival (Table 1). MULTIVARIATE ANALYSIS FOR SURVIVAL To elucidate the independent factors affecting survival, multivariate analysis using the Cox proportional hazards model was performed with the following factors: age, sex, presence of viral hepatitis, pre-olt -fetoprotein level, Child-Pugh class, size of the tumor, presence of multiple tumors, presence of bilobar tumor, microscopic vascular invasion, microscopic lymph node metastasis, histological differentiation, UICC ptnm stage based on microscopic findings, donor bone marrow infusion, and chemotherapy(transarterial Figure 2. Effect of histological differentiation in large ( 5-cm) tumors. Although survival for patients with tumors larger than 5 cm with poorly differentiated hepatocellular carcinoma (HCC) was extremely poor, patients with well- to moderately differentiated HCC with tumors larger than 5 cm had unexpectedly good survival rates. chemoembolization and intravenous chemotherapy). Multivariate analysis revealed histological differentiation and UICC ptnm stage to be the independent factors affecting survival (Table 2). Further analysis without UICC ptnm stage revealed histological differentiation and tumor size to be the independent factors affecting survival (Table 2). The outcome of patients with tumors larger than 5 cm with poorly differentiated HCC (range, 7-18 cm; median, 9.25 cm) was extremely poor (% survival within 1 year of OLT). Patients with well- to moderately differentiated HCC with tumors larger than 5 cm (range, cm; median, 6.25 cm) presented with unexpectedly good 3-year survival of 62% (Figure 2). Six of 8 patients in this group (wellto moderately differentiated tumors 5 cm) are alive withoutrecurrenceoftumorsatamedianfollow-upof38months (range,8monthsto8years).these6patientsallhadasingle nodule measuring 5.5 to 8 cm (median, 6.25 cm). Patients with small tumors did well, with 3-year survival of 82% among patients with well- to moderately differentiated HCC and 67% among patients with poorly differentiated HCC. This difference in survival within the small tumor group was not statistically significant (P=.26, log-rank test) (Figure 3). 27

4 Proportion Survival % Follow-up, d HISTOLOGICAL DIFFERENTIATION AND RECURRENCE Univariate analysis for disease-free survival was performed with the factors studied in the survival analysis. Seven factors incidental tumor, size, bilobar tumor, vascular invasion, lymph node metastasis, histological differentiation, and ptnm stage showed a statistically significant difference (P.5, log-rank test). Bone marrow infusion did not affect disease-free survival (Table 3). To characterize the independent factors affecting diseasefree survival, multivariate analysis with Cox proportional hazards model was performed similarly to the overall survival analysis. Histological differentiation and microscopic lymph node metastasis were noted as the independent factors (Table 4). Excluding lymph node metastasis, only histological differentiation remained a significant factor (Table 4). To further understand the effect of histological differentiation of HCC on prognosis, recurrence was studied according to differentiation of the tumor. Nine of 15 patients with poorly differentiated HCC died; 7 of these deaths were directly related to recurrent disease. On the other hand, 8 of 38 patients with wellto moderately differentiated HCC died; 4 of these deaths were directly related to recurrent disease (Table 5). When combined with size criteria (Table 5), 5 of 6 patients with large ( 5 cm), poorly differentiated HCC died of recurrence, although only 2 of 8 patients with large but wellto moderately differentiated HCC died of recurrence. When disease-free interval was compared in patients with recurrence based on histological differentiation, it was significantly greater in patients with well- to moderately differentiated HCC. The median disease-free interval of patients with well- to moderately differentiated HCC was 43 days, whereas that of patients with poorly differentiated HCC was 175 days (P=.6, Mann-Whitney U test) (Figure 4). COMMENT Well- to Moderately Differentiated HCC (n = 3) Poorly Differentiated HCC (n = 9) Figure 3. Effect of histological differentiation in small ( 5-cm) tumors. Patients with small tumors did well in general. Three-year survival was 82% for patients with well- to moderately differentiated hepatocellular carcinoma (HCC) and 67% in patients with poorly differentiated HCC. 82% Table 3. Disease-Free Survival According to Clinicopathologic Features* Disease-Free Survival, % Factor Patients, No. 1y 3y Age, y Sex Male Female Viral hepatitis Yes No Incidental Yes No Fetoprotein, µg/l Child-Pugh class A and B C Size, cm Multiple tumors Yes No Bilobar tumors Yes No Vascular invasion Yes No Lymph node positive Yes 2 No Histological differentiation Well and moderate Poor UICC ptnm Stage II Stage III Bone marrow infusion Yes No Chemotherapy before OLT Yes No Chemotherapy after OLT Yes No Previously, significant prognostic factors after OLT for HCC have included features such as size, number, and distribution of the tumor; nodal metastasis; vascular invasion; and TNM stage based on microscopic findings. 7,9,18-21 The role of histological differentiation of the tumor was not studied. 7,9,18-21 To date, to our knowledge, only one published study 1 from an international registry has suggested that histological differentiation is a significant and independent prognostic factor after OLT for patients with HCC. This study 1 was viewed critically by other investigators mainly because it was based P *UICC indicates Union Internationale Contre le Cancer; OLT, orthotopic liver transplantation. Kaplan-Meier method. By log-rank test for overall patient survival

5 Table 4. Cox Regression Model for Histological Differentiation and Lymph Node Metastasis and for Histological Differentiation* Factor Relative Risk (95% CI) P Analysis With Lymph Node Metastasis Histological differentiation, well and 3.6 ( ).2 moderate vs poor Microscopic lymph node metastasis, positive vs negative 12.8 ( ).9 Analysis Without Lymph Node Metastasis Histological differentiation, well and moderate vs poor 3.5 ( ).2 *CI indicates confidence interval. Relative risk of disease recurrence estimated in multivariate analysis using forward stepwise procedure in Cox regression model. on observations of different pathologists from multiple transplant centers. Our study is from the patient records of a single center where all the histological slides were collected and evaluated by a single pathologist, thereby excluding major intraobserver variability in histological differentiation. The role of OLT in the management of HCC has been controversial. Orthotopic liver transplantation has an advantage over partial resection in that it can be applied regardless of the functional reserve of the liver. Also, it is now known that there is a group of patients with HCC that presents with multicentric, metachronous hepatocarcinogenesis Total hepatectomy performed at the time of OLT allows removal of a potentially carcinogenic liver. Long-term results from earlier studies 6,8 were, nevertheless, disappointing. These series noted 2% to 3% overall 5-year survival, a high recurrence rate, and occasional long-term survival free of recurrent disease. Further studies 7,18-2,25 revealed that successful transplantation for HCC yielding results comparable to conventional partial hepatectomy can be achieved in patients with early-stage disease. These results were encouraging and suggested that OLT should be added to the therapeutic spectrum for HCC in selected patients. The most recent studies, 9,26 based on strict criteria for the size and number of tumors, observed 4-year actual survival of 75% and recurrence-free survival of 83% 9 or 5-year survival of 74%. Similarly, our overall survival of 78% and disease-free survival of 69% in 3 years for patients with tumor size of 5 cm or less reinforces the recommendation of OLT for such patients. Among patients with low-grade (well- to moderately differentiated) HCC, even better 3-year survival of 81% was observed. Furthermore, patients with large ( 5 cm) but well-differentiated tumors demonstrated 3-year survival of 62%, suggesting that these patients might also be candidates for OLT in selected circumstances. On the contrary, patients with large ( 5 cm) poorly differentiated tumors had a dismal prognosis. In our study, most of these patients presented with recurrence within a short period, and none survived more than a year. Our single-center study reaffirms the findings of the multicenter study and suggests that, aside from size and multiplicity, histological differentiation of the tumor is Disease-Free Interval, d Table 5. Histological Differentiation, Tumor Size, and Prognosis d Poor (n = 7) 43 d Differentiation Patients, No.* Alive Dead Total Differentiation Well to moderate 3 (2) 8 (5) 38 Poor 6 9 (7) 15 Tumor size and differentiation 5cm(n=39) Well to moderate 24 (2) 6 (3) 3 Poor 6 3 (2) 9 5cm(n=14) Well to moderate 6 2 (2) 8 Poor 6 (5) 6 *Number of patients with recurrent disease is shown in parentheses. One patient with adrenal metastasis died of sepsis. Well to Moderate (n = 7) P =.6 Figure 4. Comparison of disease-free intervals between well- to moderately differentiated hepatocellular carcinoma (HCC) and poorly differentiated HCC in patients with recurrence. a statistically significant factor affecting prognosis. In our multivariate analysis, histological differentiation of the tumor, an indicator of biological aggressiveness and progression, 13,27 was an independent, statistically significant factor affecting survival. Histological differentiation of the tumor was also a statistically significant factor in multivariate analysis for disease-free survival. From these results, it seems that indication for OLT should not be defined solely by size of the tumor but also with consideration of biological aggressiveness of the tumor cells as indicated by histological differentiation. Orthotopic liver transplantation for HCC has several problems that must be considered. Waiting time for a graft is often 1 year or longer, during which HCC might grow and progress. To circumvent this problem, adjunctive measures might be necessary. Effective adjuvant therapy to prevent further progression of the tumor could be applied. Because most OLT candidates have poor functional reserve of the liver, nonsurgical local ablation modalities such as thermal ablation 28,29 or percutaneous ethanol injection therapy 3 might be useful. In our series, a small number of patients received transarterial chemoembolization before OLT and chemotherapy after OLT. Although a previous study 31 suggested the beneficial ef- 29

6 fect of chemotherapy or adjuvant therapy combined with OLT for HCC, this was not clearly demonstrated in our series, probably because of the small number of patients. Further prospective studies are required. More attractive options, such as living-related liver transplantation, 32 split-liver transplantation, 33 or domino transplantation, 34,35 should be serious considerations because these measures can effectively reduce waiting time. In our experience, histological differentiation had a statistically significant effect on prognosis. Patients with small-sized HCC, regardless of histological grading, should be acceptable candidates for OLT. For relatively large tumors (slightly greater than 5 cm in diameter), OLT should not be denied solely on the basis of the size of the tumor. Because ultrasound-guided biopsy is a reliable and relatively safe method for obtaining histological information, 36 we suggest a preoperative biopsy to determine the histological differentiation should the HCC be large. Nonetheless, such an approach might have inherent, albeit small, risks of bleeding and tumor seeding. The reported rates of tumor implantation attributable to needle biopsies have varied from.5% to 5.1%, 37,38 with most of the implanted tumors occurring in the chest or the abdominal wall. Furthermore, tumor differentiation obtained through needle biopsy may not be well appreciated because of the small sample size. An alternative, therefore, may be an open or laparoscopic biopsy that can only be applied for superficial lesions. For deeper intraparenchymal lesions, an intraoperative ultrasoundguided approach seems appropriate. In conclusion, treatment protocols for HCC with OLT with or without nonsurgical modalities should consider histological grade of the tumor as a possible factor affecting survival. In addition, before we alter patient selection criteria, prospective trials are needed to confirm the impact of histological grade on the outcome of OLT for HCC. Corresponding author and reprints: Tomoaki Kato, MD, Division of Transplantation, Department of Surgery, University of Miami School of Medicine, 181 NW Ninth Ave, Highland Professional Building, Suite 511, Miami, FL REFERENCES 1. Colombo M. Hepatocellular carcinoma. J Hepatol. 1992;15: Schafer DF, Sorrell MF. Hepatocellular carcinoma. Lancet. 1999;353: Okuda K, Ohtsuki T, Obata H, et al. Natural history of hepatocellular carcinoma and prognosis in relation to treatment: study of 85 patients. Cancer. 1985;56: Venook AP. Treatment of hepatocellular carcinoma: too many options? J Clin Oncol. 1994;12: Farmer DG, Rosove MH, Shaked A, Busuttil RW. Current treatment modalities for hepatocellular carcinoma. Ann Surg. 1994;219: Penn I. Hepatic transplantation for primary and metastatic cancer of the liver. Surgery. 1991;11: Bismuth H, Chiche L, Adam R, Castaing D, Diamond T, Dennison A. Liver resection versus transplantation for hepatocellular carcinoma in cirrhotic patients. Ann Surg. 1993;218: Iwatsuki S, Gordon RD, Shaw BW Jr, Starzl TE. Role of liver transplantation in cancer therapy. Ann Surg. 1985;22: Mazzaferro V, Regalia E, Doci R, et al. Liver transplantation for the treatment of small hepatocellular carcinomas in patients with cirrhosis. N Engl J Med. 1996; 334: Klintmalm GB. Liver transplantation for hepatocellular carcinoma: a registry report of the impact of tumor characteristics on outcome. Ann Surg. 1998;228: Hermanek P, Sobin LH (International Union Against Cancer), eds. TNM Classification of Malignant Tumors. 4th ed, 2nd rev. Berlin, Germany: Springer; Kondo Y. Histologic features of hepatocellular carcinoma and allied disorders. Pathol Annu. 1985;2: Sugihara S, Nakashima O, Kojiro M, Majima Y, Tanaka M, Tanikawa K. The morphologic transition in hepatocellular carcinoma: a comparison of the individual histologic features disclosed by ultrasound-guided fine-needle biopsy with those of autopsy. Cancer. 1992;7: Edmondson HA, Steiner PE. Primary carcinoma of the liver: a study of 1 cases among 48,9 necropsies. Cancer. 1954;7: Kaplan E, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc. 1958;53: Mantel N, Haenszel W. Statistical aspects of the analysis of data from retrospective studies of disease. J Natl Cancer Inst. 1959;22: Cox D. Regression models and life tables. J R Stat Soc (B). 1972;34: McPeake JR, O Grady JG, Zaman S, et al. Liver transplantation for primary hepatocellular carcinoma: tumor size and number determine outcome. J Hepatol. 1993;18: Ringe B, Pichimayr R, Wittekind C, Tusch G. Surgical treatment of hepatocellular carcinoma: experience with liver resection and transplantation in 198 patients. World J Surg. 1991;15: Selby R, Kadry Z, Carr B, Tzakis A, Madariaga JR, Iwatsuki S. Liver transplantation for hepatocellular carcinoma. World J Surg. 1995;19: Iwatsuki S, Starzl TE, Sheahan DG, et al. Hepatic resection versus transplantation for hepatocellular carcinoma. Ann Surg. 1991;214: Yamamoto T, Kajinao K, Kudo M, Sasaki Y, Arakawa Y, Hino O. Determination of the clonal origin of multiple human hepatocellular carcinomas by cloning and polymerase chain reaction of the integrated hepatitis B virus DNA. Hepatology. 1999;29: Kumada T, Nakano S, Takeda I, et al. Patterns of recurrence after initial treatment in patients with small hepatocellular carcinoma. Hepatology. 1997;25: Eguchi A, Furuta T, Haraguchi M, Sugimachi K. A valid new approach in treating solitary new lesions after resection of hepatocellular carcinoma. J Surg Oncol. 1994;56: Otto G, Heuschen U, Hofmann WJ, Krumm G, Hinz U, Herfarth C. Survival and recurrence after liver transplantation versus liver resection for hepatocellular carcinoma: a retrospective analysis. Ann Surg. 1998;227: Llovet JM, Bruix J, Fuster J, et al. Liver transplantation for small hepatocellular carcinoma: the tumor-node-metastasis classification does not have prognostic power. Hepatology. 1998;27: Mise K, Tashiro S, Yogita S, et al. Assessment of the biological malignancy of hepatocellular carcinoma: relationship to clinicopathological factors and prognosis. Clin Cancer Res. 1998;4: Seki T, Wakabayashi M, Nakagawa T, et al. Ultrasonically guided percutaneous microwave coagulation therapy for small hepatocellular carcinoma. Cancer. 1994; 74: Curley SA, Izzo F, Delrio P, et al. Radiofrequency ablation of unresectable primary and metastatic hepatic malignancies: results in 123 patients. Ann Surg. 1999; 23: Shiina S, Tagawa K, Niwa Y, et al. Percutaneous ethanol injection therapy for hepatocellular carcinoma: results in 146 patients. AJR Am J Roentgenol. 1993; 16: Olthoff KM, Rosove MH, Shackleton CR, et al. Adjuvant chemotherapy improves survival after liver transplantation for hepatocellular carcinoma. Ann Surg. 1995; 221: Tanaka K, Uemoto S, Tokunaga Y, et al. Surgical techniques and innovations in living-related liver transplantation. Ann Surg. 1993;217: Busuttil RW, Goss JA. Split liver transplantation. Ann Surg. 1999;229: Tzakis AG, Nery JR, Raskin JB, et al. Domino liver transplantation combined with multivisceral transplantation. Arch Surg. 1997;132: Stangou AJ, Heaton ND, Rela M, Pepys MB, Hawkins PN, Williams R. Domino hepatic transplantation using the liver from a patient with familial amyloid polyneuropathy. Transplantation. 1998;65: Chapel F, Guettier C, Chastang C, et al. Needle biopsy of hepatocellular carcinoma: assessment of prognostic contribution of histologic parameters including proliferating cell nuclear antigen labeling and correlations with clinical outcomes. Cancer. 1996;77: Takamori R, Wong LL, Dang C, Wong L. Needle-tract implantation from hepatocellular cancer: is needle biopsy of the liver always necessary? Liver Transplant. 2;6: Smith EH. The hazards of fine-needle aspiration biopsy. Ultrasound Med Biol. 1984;1:

7 Invited Critique H epatocellular cancer is increasingly common in the age of chronic hepatitis B and C. Because of residual chronic hepatitis and tendency of tumor recurrence in the native liver, transplantation rather than hepatic resection is more likely to result in a successful outcome. Certain tumor descriptors exist that allow us to preoperatively stage tumors and assess the potential for recurrence following transplantation. Typically these descriptors include size, number of lesions, bilobar disease, macroscopic intrahepatic vascular invasion (into portal or hepatic venous radicals visible on contrast computed tomographic scanning), capsular invasion with local extension, and nodal or distant disease. All of this information can usually be obtained with the noninvasive high-quality imaging now available. Actual microvascular invasion can only be seen on the explanted specimen, so it is not available to factor into the preoperative decision making. However, microvascular invasion by itself can only result in a stage III designation a stage that is not a contraindication for transplantation. Transplantation for stages I-III tumors yields excellent results while outcomes for stage IVa disease are significantly worse. Placement into ptnm stages lower than III is possible using noninvasive studies, and many liver centers will not perform biopsies if the -fetoprotein level is elevated in the presence of characteristic lesions. The authors tell us that, independent of ptnm tumor stage, histological differentiation predicts outcome for those lesions larger than 5 cm. If this were true, biopsy would be indicated for every patient with lesions larger than 5 cm prior to making a decision about transplantation. But since each ptnm stage is a composite of potential variables, it would be important to do a multivariate analysis with respect to each individual variable, particularly that of major intrahepatic vascular invasion. I suspect that most of the patients with poorly differentiated tumors actually have macroscopic vascular invasion, and as such would be identified by noninvasive studies and excluded from transplantation. It is the major venous invasion, then, and not histological factors, that is the important predictor. If, on the other hand, large ( 5 cm) bilobar lesions with major venous invasion and well-differentiated histological features have a 62.5% 3-year survival rate after transplantation, but the same topography with a poorly differentiated tumor is uniformly fatal at 3 years, then a biopsy is indicated in every case of stage IVa disease. A reevaluation of the statistical analysis using not the ptnm stage but the individual variables within the staging system should reflect the importance of histological analysis as an independent variable for tumor recurrence after transplantation. Robert Rick Selby, MD Los Angeles, Calif Invited Response W e agree with Dr Selby regarding the predictive value of major vascular invasion. We also use noninvasive imaging techniques extensively and consider major vascular invasion an absolute contraindication to liver transplantation. Even so, major vascular invasion was observed in the explanted livers of 6 of our patients. Four patients had poorly differentiated tumors and 2 had well- to moderately differentiated tumors (data not included in the text). We did perform a multivariate analysis excluding ptnm stage, and major vascular invasion was not a statistically significant independent factor in either patient survival or tumor recurrence analysis, although it showed statistical significance in univariate analysis. Our findings reinforce an earlier observation of a multicenter study, 1 which suggested that the histologic grade of the tumor is an independent predictive factor of patient outcome. Our findings may not be universally applicable at this time, owing to the diversity of patient management in different transplant programs. Perhaps they can be used to guide postoperative management in most centers. However, the usefulness of our findings for patient selection will have to be decided after the value of the preoperative biopsy has been evaluated prospectively. Sumihito Tamura, MD Tomoaki Kato, MD Rajander Reddy, MD Andreas Tzakis, MD Miami, Fla 1. Klintmalm GB. Liver transplantation for hepatocellular carcinoma: a registry report of the impact of tumor characteristics on outcome. Ann Surg. 1998;228: