A, Overall survival rates of patients in the United Network of Organ Sharing database who underwent liver transplantation for benign solid liver tumors. B, Overall patient survival by diagnosis (P = .005).
Survival rates of patients with a diagnosis of benign solid liver tumors who underwent transplantation, by age (P = .005).
Sayee Sundar Alagusundaramoorthy, Valery Vilchez, Anthony Zanni, Achuthan Sourianarayanane, Erin Maynard, Malay Shah, Michael F. Daily, Luis R. Pena, Roberto Gedaly. Role of Transplantation in the Treatment of Benign Solid Tumors of the LiverA Review of the United Network of Organ Sharing Data Set. JAMA Surg. 2015;150(4):337–342. doi:10.1001/jamasurg.2014.3166
The role of orthotopic liver transplantation for the treatment of benign solid liver tumors (BSLT) is not well defined.
To analyze outcomes in the United Network of Organ Sharing data set of patients with a diagnosis of BSLT who underwent transplantation.
Design, Setting, and Participants
A retrospective analysis of the United Network of Organ Sharing data set was performed for all (N = 87 280) patients who underwent transplantation for BSLT in the United States from October 1, 1988, through January 31, 2013.
Main Outcomes and Measures
Demographics, clinicopathological characteristics, distribution of the procedures by region and state, and overall survival rates.
During the study period, 147 liver transplants (0.17%) were to treat BSLT. Sixty-two patients (42.2%) had adenomas, 29 (19.7%) had focal nodular hyperplasia, 25 (17.0%) had hemangiomas, 11 (7.5%) had hepatic epithelioid hemangioendotheliomas, and 20 (13.6%) were classified as having unknown benign tumors. The overall 1-, 3-, and 5-year survival rates were 90.9%, 85.2%, and 81.8%, respectively. Using multivariable analysis, we found that age was the only independent factor associated with patient survival. The overall 5-year survival rate for patients older than 50 years was 88% compared with 91% in younger individuals (95% CI, 148-384; P = .005). Region 3 (Alabama, Arkansas, Florida, Georgia, Louisiana, Mississippi, and Puerto Rico) contributed the maximum number (33 [22.4%]) of these transplants.
Conclusions and Relevance
Although liver transplantation cannot be considered a first-line treatment, it is a valid therapeutic option in selected patients who are not amenable to resection. Only 0.17% of the transplants in the United States are performed for this indication, with satisfying long-term results. Age was an independent predictor of patient survival. Further studies are needed to better understand the role of liver transplantation in the treatment of BSLT.
Benign solid liver tumors (BSLT) affect approximately 20% of the US population, exceeding the number of malignant tumors by a 2 to 1 ratio.1 Recent advances in imaging techniques and their widespread use have led to the more frequent and incidental detection of focal lesions in the liver.
It is well known that these lesions have different prognoses based on their characteristics; hemangiomas and focal nodular hyperplasia are typically indolent and uneventful, while hepatic adenomas can be complicated by hemorrhage, spontaneous rupture, or malignant transformation. A surgical approach to treating these tumors is only indicated in approximately 6% of patients and usually involves either resection or enucleation.2 The most common indications for surgery are presence of symptoms, uncertain diagnosis, and prevention of complications or malignant transformation.
The role of orthotopic liver transplantation and its indications have been continuously expanding. Orthotopic liver transplantation for primary and metastatic malignant neoplasms of the liver has been the subject of numerous articles.3- 6 Liver transplantation has been indicated in patients with benign tumors when lesions are highly symptomatic, diffuse in the hepatic parenchyma, or responsible for life-threatening complications or when malignant transformation cannot be ruled out. Unfortunately, published studies are limited to small case series, to our knowledge. Owing to the significant postoperative morbidity and shortage of donor livers, the indications for such aggressive therapy for benign neoplasms of the liver have yet to be defined.
The aim of this study is to evaluate the experiences of US patients from the United Network of Organ Sharing (UNOS) data set who underwent liver transplantation for BSLT.
The UNOS data set was reviewed for all patients who underwent liver transplantation in the United States from October 1988 through January 2013. Of the 87 280 total liver transplants, 147 (0.17%) were undertaken to treat BSLT.
For patients diagnosed with BSLT, the following data were collected: age, sex, type of tumor, serum creatinine level, total bilirubin level, international normalized ratio, aspartate aminotransferase level, alanine aminotransferase level, albumin level, body mass index, blood type, region where procedure was given, length of hospital stay (days), incidence of acute cellular rejection at 6 months, and overall survival rate. Donor characteristics, such as age, sex, and warm and cold ischemia times, were also recorded. Institutional review board approval was waived for this study.
Categorical variables were compared by χ2 test and continuous variables by analysis of variance (ANOVA). Overall survival and survival rates for specific groups were determined by the Kaplan-Meier test. The log-rank test was used to determine differences in survival rates. The Cox proportional hazards regression model was used to determine independent predictors of overall survival in these patients. P < .05 was considered significant. Statistical analysis was completed using SPSS software, version 20 (IBM).
During the study period, 147 liver transplants were conducted in patients with a diagnosis of BSLT. Of these patients, 62 (42.2%) had adenomas, 29 (19.7%) had focal nodular hyperplasia (FNH), 25 (17.0%) had hemangiomas, 11 (7.5%) had hepatic epithelioid hemangioendotheliomas, and 20 (13.6%) were classified as having unknown benign liver tumors at the time of transplantation. Those with a diagnosis of “unknown benign liver tumors” in the UNOS database were characterized as having lymphangiomas (1 [5.0%]), cystadenomas (2 [10.0%]), and indeterminate benign tumors (17 [85.0%]) (Table 1).
The mean (SD) age of the patients was 37.5 (18.0) years and 106 (72.1%) were female. Seventeen transplants (11.6%) were conducted in pediatric patients. Most patients were white (72.8%), followed by African American (13.6%). Sixty-three patients (42.9%) had blood group O, 58 (39.5%) blood group A, 18 (12.2%) blood group B, and 8 (5.4%) blood group AB. In 3 patients, the BSLT were found incidentally on the explanted liver.
Mean (SD) values at transplantation were the following: for serum creatinine level, 1.2 (1.5) mg/dL (to convert to micromoles per liter, multiply by 88.4); total bilirubin level, 2.5 (5.8) mg/dL (to convert to micromoles per liter, multiply by 17.104); international normalized ratio, 0.9 (0.7); albumin level, 3.4 (1.1) g/dL (to convert to grams per liter, multiply by 10); alanine aminotransferase level, 51.6 (225.3) U/L (to convert to microkatals per liter, multiply by 0.0167; and aspartate aminotransferase level, 18.2 (30.2) U/L (to convert to microkatals per liter, multiply by 0.0167). Forty patients from this cohort had a total bilirubin level greater than 1.5 mg/dL. In most cases, deceased donors were used (136 patients [92.5%]), and only 11 (7.5%) living-donor liver transplants were performed. The mean (SD) donor age at the time of transplantation was 35.4 (19.0) years. The mean (SD) cold ischemic time was 7.6 (4.7) hours, and the warm ischemic time was 45.2 (24.2) minutes. The mean (SD) body mass index (calculated as weight in kilograms divided by height in meters squared) was 24.8 (7.3). Nearly 47.6% of the patients had upper abdominal surgery at the same time as or before the transplantation procedure. The mean (SD) duration of posttransplant hospitalization was 15.8 (19.9) days (Table 2).
Of 93 patients who received transplants after 2002 (63.3%), when the Model for End-Stage Liver Disease (MELD) score was implemented, 65 (69.9%) received transplants with MELD-score exceptions. The mean (SD) MELD score at transplantation was 11.8 (6.4). Twenty-nine patients had a MELD score higher than 15 at the time of transplantation. Of 28 patients who underwent transplantation without exception during the MELD era (after 2002), 16 (57.1%) had MELD scores higher than 15. We compared age, sex, body mass index, weight, creatinine level, total bilirubin level, length of stay, and survival rates among patients who underwent transplantation before and after their MELD score was implemented. We found that patients in the pre-MELD era had significantly higher total bilirubin levels than those in the post-MELD era (mean pre-MELD bilirubin level, 3.7 mg/dL vs post-MELD level, 1.8 mg/dL).
Twenty patients (13.6%) in this cohort were treated for acute cellular rejection within 6 months. A total of 132 patients (89.8%) had a functioning graft during the mean follow-up time of 79.2 months.
Region 2 (Delaware, District of Columbia, Maryland, New Jersey, Pennsylvania, West Virginia, and Northern Virginia), Region 3 (Alabama, Arkansas, Florida, Georgia, Louisiana, Mississippi, and Puerto Rico), Region 4 (Oklahoma and Texas), and Region 5 (Arizona, California, Nevada, New Mexico, and Utah) contributed 92 (62.6%) of these transplants (Table 3). Thirty-three (22.4%) were conducted in Region 3. California and Texas had more procedures (16 each, or 10.9%). Florida and Pennsylvania conducted 13 (8.8%) and 11 (7.5%), respectively.
The overall survival rates at 1, 3, and 5 years were 90.9%, 85.2%, and 81.8%, respectively (Figure 1A). The overall graft survival rates at 1, 3, and 5 years was 85.6%, 79.7%, and 75.6%, respectively.
In patients who were diagnosed with adenomas, the overall survival rates at 1, 3, and 5 years were 98.4%, 94.1%, and 91.6%, respectively. In those diagnosed with FNH, the overall survival rates at 1, 3, and 5 years were 81.8%, 76.3%, and 68.7%, respectively. Patients diagnosed with hepatic hemangioma who underwent liver transplantation had overall survival rates at 1, 3, and 5 years of 87.8%, 81.5%, and 74.8%, respectively (Figure 1B).
Interestingly, among patients who underwent transplantation for FNH, there was a trend toward decreased survival compared with other diagnoses (Figure 1B). There were no statistical differences in survival rates among patients with BSLT by diagnosis (adenomas, FNH, hepatic epithelioid hemangioendothelioma, hemangioma, and others).
On multivariable analysis and after controlling by age, sex, and MELD scores higher than 15, we found that age was the only independent factor associated with survival in patients who underwent transplantation for BSLT (95% CI, 1.004-1.078; P = .03).
Because age was a significant predictor of outcome in this patient population, we divided the sample into quartiles. Patients who underwent transplantation and were older than 50 years were considered to be above the 75th percentile. The overall 5-year survival rate among older patients was 88.0% compared with 91.0% in younger individuals (95% CI, 148-384; P = .005) (Figure 2). This difference was more significant 10 to 15 years after the procedure was performed.
Benign solid liver tumors are being diagnosed frequently among the general population with the increased use of various imaging techniques. Liver resection is only indicated in a few cases in the presence of complications or when malignancy cannot be ruled out. Interestingly, the number of liver transplants undertaken for diffuse unresectable benign tumors has been slowly increasing worldwide.
In a review of the US experience, resection for benign liver lesions represented about 1% to 3% of all hepatectomies.7 In an analysis of 1222 hepatectomies, Poon et al8 reported that liver resection was performed for benign liver lesions in 15.4% of patients. Similarly, Jarnagin et al9 reported that, at their institution, 9% of 1803 hepatectomies were for BSLT.
In a retrospective study of 3239 liver transplant recipients at the University of Pittsburgh from 1981 through 1993, Tepetes et al10 reported that 12 transplants (0.37%) were for unresectable BSLT. Of these cases, 6 had adenomas, 2 mesenchymal hamartomas, 1 massive hepatic lymphangiomatosis, 1 hilar fibrous angiodysplasia, 1 FNH, and 1 hemangioma. The median duration of survival for the 9 patients who survived the perioperative period was 88 months. Based on their experience, the authors concluded that patients with severe symptoms caused by benign hepatic tumors that are not resectable can be treated by total hepatectomy and orthotopic liver transplantation, with the expectation of good long-term results. Ercolani et al,3 using the European registry, reported that 1.2% of liver transplants conducted in Europe are for benign liver tumors.
In our analysis of the UNOS data set from October 1988 to January 2013, we were able to find 147 (0.2%) liver transplants that were undertaken to treat BSLT with good long-term results. We found overall survival rates at 1, 3, and 5 years of 90.9%, 85.2%, and 81.8%, respectively. This is similar to the reported rates for patients who received transplants for other benign diseases.11,12
As expected, most of the patients were young (mean overall age, 35.4 years) and female (associated with a higher prevalence of adenomas and FNH) and displayed mostly normal results of liver function tests. Most of these patients had MELD scores that were lower than 15. Interestingly, 40 patients had abnormal total bilirubin levels, suggesting that several had some degree of underlying liver disease. In most of these cases, we were not able to find a clear diagnosis of cirrhosis or a disease that would cause abnormal liver function because the secondary diagnosis, in some instances, was not well documented in the UNOS database.
Hepatic adenomas have been reported to be a frequent complication of metabolic diseases, including type I glycogen storage disease, galactosemia, and tyrosinemia.4 Hepatic adenomas are usually asymptomatic but can be associated with right upper-quadrant pain that is sometimes related to spontaneous hemorrhage. This discharge usually occurs in lesions that are larger than 5 cm and have an associated risk of death of 8%. The risk of hemorrhage is directly attributed to the size of the largest nodule rather than the number of nodules. There is a reported13 5% risk of malignancy in hepatic adenomas, with 28% overall risk of bleeding. In general, accepted indications for surgical resection of these lesions include symptomatic tumors, diagnostic uncertainty, and tumors larger than 5 cm in diameter.14,15 Usually, patients with hepatic adenomas or liver adenomatosis maintain normal liver function without portal hypertension, ascites, or other complications related to cirrhosis. Based on our analysis, the most common benign solid tumor type diagnosed at the time of transplantation was adenoma, with 62 cases (42.2%) and overall survival rates at 1, 3, and 5 years of 98.4%, 94.1%, and 91.6%, respectively.
The prevalence of FNH has been reported to be around 0.6% to 3.0% in the general population.10 Some benign hepatic neoplasms, such as FNH, adenoma, and nodular regenerative hyperplasia, are frequently associated with conditions that cause alteration in the hepatic perfusion, including intrahepatic portal vein agenesis. These lesions have also been reported16,17 to be associated with the use of immunosuppressive agents, especially when azathioprine is part of the regimen, and with living donor liver transplantation in cases of small-for-size syndrome. In this cohort, 29 of the patients had FNH, with overall survival rates after orthotopic liver transplantation at 1, 3, and 5 years of 81.8%, 76.3%, and 68.7%, respectively. We found that FNH survival rates were among the least favorable compared with other benign tumors in our series; however, this difference was not statistically significant.
Hepatic hemangiomas are the most common benign liver tumor, with prevalence in imaging studies and autopsy findings of up to 7%.5 In this cohort, 25 patients underwent liver transplantation after the diagnosis of hemangioma, with overall survival rates at 1, 3, and 5 years of 87.8%, 81.5%, and 74.8%, respectively. Spontaneous hemorrhage is infrequent and usually associated with accidental rupture during biopsy or trauma. Like most other benign tumors, hemangiomas are asymptomatic and usually managed conservatively. In a few cases, resection may be needed, usually as a consequence of intractable pain or inability to establish a diagnosis, in cases of a rapidly growing lesion for which malignancy cannot be excluded, and in rare cases of Kasabach-Merritt syndrome.
Infantile hepatic hemangioma is a clinically different entity that presents as diffuse disease in 10% to 20% of diagnosed infants, sometimes refractory to medical therapy, and is associated with life-threatening complications. Interestingly, in our pediatric cohort of 17 patients, 5 received transplants for adenomas, 5 for FNH, and 2 for hepatic epithelioid hemangioendothelioma; the remaining 5 had an indeterminate diagnosis. Interestingly, none of the pediatric patients in our series underwent liver transplantation for infantile hepatic hemangioma.
Hepatic epithelioid hemangioendothelioma is a rare soft-tissue vascular tumor with an intermediate clinical course between benign hemangioma and malignant angiosarcoma. We identified only 11 patients with this diagnosis who received transplants; the overall survival rates at 1, 3, and 5 years were 90.0%, 78.8%, and 78.8%, respectively. These results are similar to those of a study of patients in the European registry in which overall survival was 83% and 78% at 5 and 10 years, respectively, in 59 patients who received transplants for this indication.10 In 2009, Cardinal et al18 reported the results of a study at the University of Pittsburgh from 1976 to 2007 in which 25 patients received a diagnosis of hepatic epithelioid hemangioendotheliomas. Of these patients, 17 underwent liver transplantation, 4 underwent transcatheter arterial chemoembolization alone, 2 underwent transcatheter arterial chemoembolization followed by liver transplantation, and 2 underwent liver resection. The mean total overall survival was 167 months and, in the liver transplantation cohort, 172 months. The authors concluded that liver transplantation is indicated in patients with advanced local disease without extrahepatic involvement or comorbidities that preclude liver transplantation. Other investigators6 have reported the incidence of extrahepatic disease of the bones, lungs, and lymph nodes to be as high as 36% of the patients studied. Hepatectomy is still considered the first line of treatment for patients with resectable hepatic epithelioid hemangioendotheliomas.
The UNOS database recorded 1 patient with a diagnosis of lymphangioma, 2 with cystadenomas, and 17 with benign tumors of unknown origin. Cystadenomas represent 5% of all cysts based on an autopsy series.19 The mainstay of treatment is complete resection, either by enucleation or formal hepatectomy. Lymphangiomas are also rare tumors that are characterized by the proliferation of lymphatic tissue, which may cause organ dysfunction by compression or replacement of adjacent parenchyma. Only 4 cases of hepatic lymphangioma that received transplants have been reported.18 All cases were reported in women; there was a 50% rate of recurrence.
We also analyzed the distribution of cases by the diagnosis and region where the procedure was performed. We found that regions 2, 3, 4, and 5 contributed to 92 (62.6%) of the total transplants. Region 3, which includes Alabama, Arkansas, Florida, Georgia, Louisiana, Mississippi, and Puerto Rico, reported 33 (22.4%) of these cases. When we examined the distribution by state, we found that California and Texas conducted 22% of the total transplants. Florida and Pennsylvania conducted 8.8% and 7.5%, respectively. In our multivariable analysis, age was found to be the only independent predictor of survival (P = .03) in patients who underwent transplantation for BSLT. Because age was strongly associated with outcomes in this analysis, we then divided the cohort into quartiles of age. Patients older than 50 years were considered to be above the 75th percentile. The overall 5-year survival rate of patients older than 50 years was 88% compared with 91% in younger individuals; this difference is even more significant 10 or 15 years after transplantation, as seen in Figure 2.
The strength of our study comes not only from the large number of patients but also from its clinically robust and uniform definitions of patient characteristics and events. However, its limitations are mostly related to the use of the UNOS database, which was not developed to study this specific patient population. For that reason, key variables are not captured, such as specific histologic data (eg, tumor size, location, and number or the percentage of liver parenchyma involved with tumor); intraoperative information, such as operative time, blood product use, and type of surgical technique (piggyback20 vs conventional); and postoperative morbidity.
Although liver transplantation cannot be considered a first-line treatment for patients with a diagnosis of BSLT, it is a valid therapeutic option in selected patients who are not amenable to resection. Age older than 50 years was associated with worse patient survival. Most of the cases were managed in highly populated states, such as California, Texas, Florida, and Pennsylvania. Future studies are needed to determine the real role of liver transplantation in the treatment of BSLT.
Accepted for Publication: July 9, 2014.
Corresponding Author: Roberto Gedaly, MD, Department of Surgery, College of Medicine, University of Kentucky, 800 Rose St, Room C453, Lexington, KY 40536 (firstname.lastname@example.org).
Published Online: February 25, 2015. doi:10.1001/jamasurg.2014.3166.
Author Contributions: Dr Gedaly had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Alagusundaramoorthy, Daily, Pena, Gedaly.
Acquisition, analysis, or interpretation of data: Alagusundaramoorthy, Vilchez, Zanni, Sourianarayanane, Maynard, Shah, Gedaly.
Drafting of the manuscript: Alagusundaramoorthy, Vilchez, Zanni, Sourianarayanane, Shah, Daily, Gedaly.
Critical revision of the manuscript for important intellectual content: Vilchez, Sourianarayanane, Maynard, Pena, Gedaly.
Statistical analysis: Alagusundaramoorthy, Vilchez, Sourianarayanane, Shah, Gedaly.
Obtained funding: Gedaly.
Administrative, technical, or material support: Alagusundaramoorthy, Vilchez, Gedaly.
Study supervision: Shah, Pena, Gedaly.
Conflict of Interest Disclosures: None reported.