SEARCH

SEARCH BY CITATION

Keywords:

  • hepatocellular carcinoma;
  • liver transplantation;
  • disparities;
  • race;
  • insurance

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. FUNDING SOURCES
  7. REFERENCES

BACKGROUND:

Prior studies have demonstrated that among patients with hepatocellular carcinoma (HCC), African Americans (AAs) and Asian/Pacific Islanders (APIs) are substantially less likely to undergo liver transplantation (LT) compared with whites. The authors examined whether disparities in the receipt of LT among LT-eligible HCC patients changed over a 10-year time period, and whether the disparities might be explained by sociodemographic or clinical factors.

METHODS:

The National Cancer Data Base, a national hospital-based cancer registry, was used to study 7707 adults with small (≤ 5 cm), nonmetastatic HCC diagnosed between 1998 and 2007. Racial/ethnic patterns in the use of LT were compared during 2 periods of 5 years each: 1998 through 2002 (n = 2412 patients) and 2003 through 2007 (n = 5295 patients). Data regarding comorbid medical conditions were only available during the later time period.

RESULTS:

Large and persistent racial/ethnic differences in the probability of receiving LT were observed. Compared with whites, hazard ratios (HRs) and associated 95% confidence intervals (95% CIs) for receiving LT from 1998 through 2002 were 0.64 (95% CI, 0.46-0.89) for AA patients, 1.01 (95% CI, 0.79-1.29) for Hispanic patients, and 0.52 (95% CI, 0.39-0.68) for API patients. Analogous results for 2003 through 2007 were 0.64 (95% CI, 0.54-0.76) for AA patients, 0.86 (95% CI, 0.75-0.99) for Hispanic patients, and 0.58 (95% CI, 0.49-0.69) for API patients. AA patients were less likely than whites to undergo any form of surgery, and API patients were more likely than whites to undergo surgical resection. Adjustment for sociodemographic and clinical factors produced only small changes in these HRs.

CONCLUSIONS:

Between 1998 and 2007, there were large and persistent racial/ethnic disparities noted in the receipt of LT among patients with HCC. These disparities were not explained by sociodemographic or clinical factors. Cancer 2011;. © 2011 American Cancer Society.

Hepatocellular carcinoma (HCC) is an aggressive disease with a poor prognosis. In the United States, the overall 5-year relative survival rate for HCC is 13.8%.1 An estimated 24,120 new cases of liver cancer were expected to occur in the United States during 2010. Greater than 80% of these cases are HCC. An estimated 18,910 liver cancer deaths were expected in 2010.2 In the past 30 years, the incidence and mortality rates for HCC in the United States have more than doubled, with the increase attributed to the rising rates of hepatitis C (HCV) infection in the United States, and migration from hepatitis B (HBV)-endemic areas.3 Liver transplantation (LT) is an effective treatment for some patients with early HCC. Among patients who met the Milan criteria for eligibility for LT (defined as a single lesion measuring ≤ 5 cm or up to 3 separate lesions, none larger than 3 cm; no evidence of macroscopic vascular invasion; and no regional lymph node or distant metastases) (hereafter referred to as “LT-eligible HCC”) and who underwent LT, the 4-year survival rate was found to be 75%,4 which is similar to outcomes in patients who underwent LT for nonmalignant indications.5 In 2002, the Organ Procurement and Transplantation Network (OPTN) began using the Model for End-Stage Liver Disease (MELD) allocation system, which gave priority for LT to HCC patients who met the Milan criteria. Implementation of the MELD system led to a 6-fold increase in the percentage of patients with HCC as their indication for LT.6

Several studies have demonstrated that among patients with LT-eligible HCC, African Americans (AAs) and Asian/Pacific Islanders (APIs) are substantially less likely to undergo LT compared with whites. In a recent study using detailed clinical data on 462 patients seen between 2002 and 2008 at a large university medical center, Yu et al reported that AA patients with HCC were 30 times less likely than whites to undergo LT.7 Two earlier studies using national data from the Surveillance, Epidemiology, and End Results (SEER) database found that AA patients were 50% to 58% less likely to undergo LT for LT-eligible HCC than whites.8, 9 One of these studies, using data from 1998 through 2002, also found that API patients with LT-eligible HCC were approximately half as likely as whites to undergo LT.8 Another study using data from the SEER and OPTN databases found that among HCC patients who underwent LT, AA patients had the worst graft survival and overall survival.10

Racial disparities in the use of LT extend well beyond patients with HCC. In a national study using data from 1998 through 2003, AA patients with end-stage liver disease were less likely than whites to be listed for LT and once listed, less likely to undergo LT. The overall rate of LT was 3 times as high among whites as among AA patients.11 Among patients diagnosed before the introduction of the MELD system, AA patients were less likely than white patients to undergo LT and more likely to die or become too sick for LT within 3 years of registering on the waiting list. After the introduction of the MELD score, AA race was no longer associated with the receipt of LT or an increased likelihood of death while on the waiting list.12 Because this study included patients with all forms of end-stage liver disease and only patients who were registered on the LT waiting list, we conducted a study focused on patients with HCC, irrespective of registration on the LT waiting list, using data from the National Cancer Data Base (NCDB). We examined whether disparities in the use of LT for HCC changed over a 10-year time period, and whether any disparities noted might be explained by differences in neighborhood socioeconomic status (SES), insurance status, geographic region, tumor size, or comorbid medical conditions.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. FUNDING SOURCES
  7. REFERENCES

The NCDB is a national hospital-based cancer registry supported by the Commission on Cancer of the American College of Surgeons and the American Cancer Society. Approximately 1430 hospitals currently report to the NCDB, which captures approximately 70% of newly diagnosed cancers in the United States. NCDB data for the current study were collected according the Registry Operations and Data Standards (ROADS) manual13 (1998-2002), the Facility Oncology Registry Data Standards (FORDS) manual14 (2003-2007), the fifth and sixth editions of the American Joint Committee on Cancer (AJCC) staging manual (denoted as AJCC 5 and AJCC 6, respectively),15, 16 and the third edition of the International Classification of Diseases for Oncology (ICD-O-3).17

We initially selected all adult patients aged ≥18 years with invasive primary HCC (ICD-O-3 site code C220, histology code 8170) who were diagnosed between 1998 and 2007 and who received their first course of treatment at a reporting facility that participated in the NCDB continuously from 1998 to 2007. We only included patients with microscopically confirmed diagnoses for whom HCC was their first malignant primary tumor (n = 43,377). We then excluded 19,093 patients with missing data regarding AJCC stage of disease (n = 10,232) or tumor size (n = 8861), leaving 24,284 patients. Because the goal of the current study was to evaluate the use of LT among LT-eligible patients with HCC, we first classified patients as LT-eligible or LT-ineligible based on AJCC stage of disease and tumor size. For patients diagnosed under AJCC 5 (1998-2002), those with stage I or stages II or IIIA disease with a tumor size ≤ 5.0 cm were classified as LT-eligible, and those with stages II or IIIA disease with a tumor size > 5.0 cm or stages IIIB or IV disease were classified as LT-ineligible. For patients diagnosed under AJCC 6 (2003-2007), those with stage I disease with a tumor size ≤ 5.0 cm or stage II disease were classified as LT-eligible, and those with stage I disease with a tumor size > 5.0 cm or stages III or IV disease were classified as LT-ineligible. Thus, among the 24,284 patients noted above, 10,066 (41.5%) were classified as LT-eligible and 14,218 (58.6%) were classified as LT-ineligible. After preliminary analyses comparing the percentage of LT-eligible patients by race/ethnicity, only LT-eligible patients were included in subsequent analyses.

Among LT-eligible patients, we further excluded 2359 cases with missing data concerning the following variables: vital status or survival time (n = 385); insurance status (n = 345); neighborhood education or income level (n = 684); age or sex (n = 4); race or ethnicity (n = 595); or type of surgery (n = 219). We also excluded racial/ethnic groups other than white, AA, Hispanic, or API (n = 127). Thus, a total of 7707 cases were available for analysis (2412 diagnosed during 1998-2002 and 5295 diagnosed during 2003-2007). The Morehouse School of Medicine Institutional Review Board (IRB) determined that this and other studies using deidentified NCDB data were exempt from full IRB review.

Based on detailed codes for surgical treatment of the primary disease site, surgery was classified as LT, surgical resection, ablation, or no surgery. Beginning in 2003, the NCDB began including data regarding comorbid medical conditions (up to 6 per patient). On the basis of secondary diagnosis fields, we classified comorbid conditions using the methodology described by Elixhauser et al, which assesses the presence or absence of 30 separate conditions.18 Since 1996, the NCDB has collected information regarding insurance status at the time of diagnosis. Insurance status was classified as private insurance, no insurance, Medicaid insurance, Medicare insurance, or other public insurance. Using ZIP code-level data from the 2000 US Census, neighborhood education level was measured by the percentage of residents in the patient's ZIP code without a high school diploma, and neighborhood income level was measured by the median household income in the patient's ZIP code. Both variables were coded using quartiles based on national distributions.

To examine trends in racial/ethnic patterns in surgical treatment, we used data from 1998 through 2007, which we divided into two 5-year periods: 1998 through 2002 and 2003 through 2007. To examine predictors of surgical treatment in a multivariate model that included comorbid medical conditions, we used data from cases diagnosed between 2003 and 2007. Preliminary analyses revealed that among LT-eligible patients, AAs had a higher death rate after diagnosis with HCC whereas APIs had a lower death rate and thus, the different racial/ethnic groups did not have the same time opportunity to receive surgery. To correct for this potential bias, we analyzed treatment for HCC using Cox proportional hazards models in which we modeled the hazard rate of receiving a specific treatment, and patients were censored on the date of death or last contact, whichever came first. To examine predictors of 5-year survival, we used Cox proportional hazards regression analysis and data from cases diagnosed between 1998 and 2002 and followed through 2007. Cases were coded as alive, dead, or lost to follow-up, and cause of death was not available. Patients who were lost to follow-up before 5 years were censored as of the date of their last contact, and patients who were followed for > 5 years were censored at 5 years. Approximately 9.5% of patients were lost to follow-up after 5 years. All statistical analyses were performed using SAS statistical software (version 9.2; SAS Institute Inc, Cary, NC). All statistical analyses used 2-sided P values and an α level of .05 was used to test for statistical significance.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. FUNDING SOURCES
  7. REFERENCES

In both the earlier and later years of the study period, there were major racial/ethnic differences in surgical treatment noted among LT-eligible patients with HCC (Table 1). In the earlier period, AA and API patients were both less likely to undergo LT, with AAs less likely than whites to undergo any form of surgery and APIs more likely than whites to undergo surgical resection. As noted earlier, there were substantial differences in death rates during follow-up; API patients had the lowest death rate, whereas AA patients had the highest death rate. In the proportional hazards analyses that accounted for different death rates during follow-up, AA patients had a 36% lower rate of undergoing LT, and API patients had a 48% lower rate. In the later period, AA, Hispanic, and API patients were all less likely to undergo LT. Again, the difference between white and AA patients appeared to mainly be because of an increased probability of no surgery among AA patients, and the difference between white and API patients appeared to mainly be because of an increased probability of surgical resection among API patients. The difference between white and Hispanic patients appeared to mainly be because of an increased probability of ablation among Hispanic patients. In the proportional hazards analyses, AA patients had a 36% lower rate of undergoing LT, Hispanic patients had a 14% lower rate, and API patients had a 42% lower rate.

Table 1. Racial/Ethnic Differences in Surgical Treatment Among LT-Eligible Patients With Hepatocellular Carcinoma, 1998-2007 (n = 7707)
Treatment/Outcome, by Diagnosis YearRace/EthnicityPa
WhiteAfrican AmericanHispanicAsian/Pacific Islander
  • Abbreviations: 95% CI, 95% confidence interval; HR, hazard ratio; LT, liver transplantation.

  • a

    For chi-square test of association between race/ethnicity and the treatment/outcome.

1998-2002 (n = 2412)     
Probability of receiving treatment, %     
 LT30.214.628.519.2<.0001
 Surgical resection21.126.414.832.9<.0001
 Ablation14.716.514.817.5.61
 No surgery34.142.541.930.5<.01
 Death rate during follow-up (per 1000 person-y)18.924.018.113.0<.0001
HR for receipt of treatment (95% CI)     
 LT1.00 (referent)0.64 (0.46-0.89)1.01 (0.79-1.29)0.52 (0.39-0.68)<.0001
 Surgical resection1.00 (referent)1.52 (1.17-1.96)0.73 (0.52-1.01)1.31 (1.04-1.64)<.001
 Ablation1.00 (referent)1.30 (0.94-1.80)1.02 (0.73-1.43)1.04 (0.77-1.41).48
 No surgery1.00 (referent)1.37 (1.12-1.69)1.24 (1.01-1.52)0.81 (0.65-1.01)<.001
2003-2007 (n = 5295)     
Probability of receiving treatment, %     
 LT38.223.633.724.3<.0001
 Surgical resection17.621.613.932.3<.0001
 Ablation14.916.618.919.6<.01
 No surgery29.438.233.423.8<.0001
 Death rate during follow-up (per 1000 person-y)25.232.922.516.1<.0001
HR for receipt of treatment (95% CI)     
 LT1.00 (referent)0.64 (0.54-0.76)0.86 (0.75-0.99)0.58 (0.49-0.69)<.0001
 Surgical resection1.00 (referent)1.26 (1.05-1.52)0.78 (0.63-0.97)1.70 (1.45-2.01)<.0001
 Ablation1.00 (referent)1.15 (0.93-1.43)1.25 (1.03-1.52)1.22 (0.99-1.49).05
 No surgery1.00 (referent)1.34 (1.16-1.54)1.13 (0.97-1.30)0.77 (0.64-0.92)<.0001

Table 2 demonstrates that there were also large racial/ethnic differences in the demographic, tumor, and clinical characteristics among LT-eligible patients with HCC. Compared with white patients, AA and Hispanic patients were younger, whereas API patients were older; all nonwhite groups had a higher proportion of females. AA and Hispanic patients were more likely than white patients to reside in neighborhoods with low education and income levels; the opposite was observed for API patients. Compared with white patients, AA and Hispanic patients were more likely to be uninsured or have Medicaid insurance, whereas API patients were more likely to have private insurance. Geographic comparisons indicated that compared with white patients, AA patients were more likely to reside in the South, and Hispanic and API patients were more likely to reside in the West. Although the racial/ethnic differences were not found to be statistically significant, AA patients had the largest mean tumor size. For 22 of 27 comorbid medical conditions, either AA or Hispanic patients had the highest prevalence.

Table 2. Characteristics of LT-Eligible Patients With Hepatocellular Carcinoma, 2003-2007, by Race/Ethnicity (n = 5295)
CharacteristicRace/EthnicityPa
WhiteAfrican AmericanHispanicAsian/Pacific Islander
  • Abbreviations: AIDS, acquired immunodeficiency syndrome; LT, liver transplantation.

  • a

    For chi-square test of association between race/ethnicity and the characteristic.

Mean age, y61.557.460.362.0<.0001
Sex, %     
 Male77.473.772.371.2<.001
 Female22.626.327.728.9 
Percentage of residents without high school degree, %     
 <14.036.914.516.941.8 
 14.0−19.924.115.213.620.5<.0001
 20.0−28.924.529.823.022.2
 ≥29.014.540.546.515.6 
Median household income, %     
 ≥$46,00041.818.428.758.4 
 $35,000-$45,99929.121.624.324.5<.0001
 $30,000-$34,99917.421.521.310.3 
 <$30,00011.738.625.76.8 
Insurance status, %     
 Private44.841.035.551.4 
 Uninsured2.66.34.81.9 
 Medicaid9.019.819.412.1<.0001
 Medicare43.131.939.234.4 
 Other government0.61.01.20.2 
Geographic region, %     
 Northeast19.717.715.715.9 
 Midwest23.426.310.98.7<.0001
 South40.248.239.918.0 
 West16.67.733.657.3 
 Mean tumor size, cm4.14.73.94.3.10
Comorbid medical conditions     
 No comorbidities listed31.629.232.037.0.03
 Congestive heart failure3.13.12.21.4.09
 Cardiac arrhythmias4.41.62.83.3<.01
 Valvular heart disease1.80.30.70.9<.001
 Pulmonary circulation disorders0.70.31.00.4.33
 Peripheral vascular disorders1.00.80.40.0.07
 Hypertension20.028.920.121.2<.0001
 Paralysis0.00.20.00.0.38
 Other neurological disorders0.90.50.40.4.27
 Chronic pulmonary disease6.97.43.42.3<.0001
Comorbid medical conditions     
 Diabetes, uncomplicated16.618.719.813.3.01
 Diabetes, complicated1.10.32.20.7<.01
 Hypothyroidism0.00.00.00.0
 Renal failure1.32.42.90.5.001
 Liver disease47.548.951.247.7.36
 Peptic ulcer disease excluding bleeding0.40.70.70.7.67
 AIDS0.22.10.70.0<.0001
 Rheumatoid arthritis/collagen vascular diseases0.30.20.70.4.32
 Coagulopathy7.36.57.64.7.12
 Obesity1.00.71.00.2.22
 Weight loss1.01.30.70.2.17
 Fluid and electrolyte disorders6.76.87.65.1.34
 Blood loss anemia0.50.20.60.4.63
 Deficiency anemias3.12.93.22.3.72
 Alcohol abuse7.57.98.71.2<.0001
 Drug abuse1.44.71.60.4<.0001
 Psychoses0.61.30.40.0.03
 Depression2.02.60.70.9.02

Among LT-eligible patients, even after adjustment for age, sex, neighborhood SES, insurance status, geographic region, tumor size, and comorbid medical conditions, AA and API patients were significantly less likely to undergo LT when compared with white patients (Table 3). Compared with white patients, AA patients were 38% less likely to undergo LT, and API patients were 33% less likely. In addition to race/ethnicity, all of the other factors except sex and neighborhood SES were found to be significant predictors of undergoing LT. Each 1-year increase in age was associated with a 7% reduction in the rate of undergoing LT. Similarly, each 1-cm increase in tumor size was associated with an 8% reduction in the rate of undergoing LT. Patients who were uninsured were 59% less likely to undergo LT. Compared with patients residing in the Northeast, those residing in the Midwest, South, or West were approximately twice as likely to undergo LT. Several comorbid medical conditions were found to be associated with an increased rate of undergoing LT, whereas several were associated with a decreased rate.

Table 3. Multivariate Analysis of Racial/Ethnic Differences in LT Among LT-Eligible Patients With Hepatocellular Carcinoma, 2003-2007 (n = 5295)
CharacteristicMultivariate HR for LT (95% CI)
  • Abbreviations: 95% CI, 95% confidence interval; HR, hazard ratio; LT, liver transplantation.

  • a

    Hypothyroidism, paralysis, and acquired immunodeficiency syndrome could not be included in the model because there were no cases, either among all LT-eligible patients or among patients receiving LT.

Race/ethnicity 
 White1.00 (referent)
 African American0.62 (0.52-0.74)
 Hispanic0.88 (0.76-1.02)
 Asian/Pacific Islander0.67 (0.56-0.81)
 Age (1-y increase)0.97 (0.96-0.97)
Sex 
 Male1.00 (referent)
 Female0.98 (0.87-1.10)
Percentage of residents without high school degree, % 
 <14.01.00 (referent)
 14.0−19.90.99 (0.86-1.13)
 20.0−28.91.05 (0.90-1.23)
 ≥29.00.89 (0.74-1.07)
Median household income 
 ≥$46,0001.00 (referent)
 $35,000−$45,9991.02 (0.89-1.16)
 $30,000−$34,9990.98 (0.83-1.15)
 <$30,0000.94 (0.78-1.15)
Insurance status 
 Private1.00 (referent)
 Uninsured0.41 (0.28-0.60)
 Medicaid0.86 (0.73-1.00)
 Medicare0.89 (0.79-1.00)
 Other government0.86 (0.47-1.57)
Geographic region, % 
 Northeast1.00 (referent)
 Midwest1.66 (1.42-1.95)
 South2.32 (2.01-2.69)
 West1.61 (1.37-1.90)
 Tumor size (1-cm increase)0.92 (0.89-0.94)
Comorbid medical conditionsa 
 No comorbidities listed1.00 (referent)
 Congestive heart failure0.76 (0.49-1.18)
 Cardiac arrhythmias1.09 (0.79-1.49)
 Valvular heart disease0.83 (0.49-1.40)
 Pulmonary circulation disorders0.75 (0.44-1.28)
Comorbid medical conditionsa 
 Peripheral vascular disorders0.52 (0.21-1.25)
 Hypertension0.87 (0.76-1.00)
 Other neurological disorders1.32 (0.78-2.25)
 Chronic pulmonary disease0.66 (0.51-0.87)
 Diabetes, uncomplicated1.08 (0.95-1.23)
 Diabetes, complicated1.09 (0.70-1.71)
 Renal failure1.23 (0.83-1.83)
 Liver disease1.37 (1.24-1.52)
 Peptic ulcer disease excluding bleeding1.20 (0.50-2.90)
 Rheumatoid arthritis/collagen vascular diseases1.85 (0.83-4.15)
 Coagulopathy1.17 (0.99-1.37)
 Obesity0.68 (0.40-1.13)
 Weight loss1.14 (0.71-1.83)
 Fluid and electrolyte disorders1.17 (0.98-1.39)
 Blood loss anemia1.78 (0.73-4.33)
 Deficiency anemias1.47 (1.15-1.88)
 Alcohol abuse0.80 (0.66-0.97)
 Drug abuse1.23 (0.83-1.82)
 Psychoses0.71 (0.32-1.60)
 Depression0.97 (0.69-1.36)

Table 4 shows the impact of several factors on 5-year survival among LT-eligible patients with HCC. AA patients were 36% more likely to die within 5 years of diagnosis, after adjustment for age and sex. After additional adjustment for undergoing LT, there was no longer any significant difference with regard to survival noted between AA and white patients. Patients who underwent LT were 73% less likely to die within 5 years of diagnosis. The results of the current study suggest that racial differences in the receipt of LT accounted for 72% of the survival difference between AA and white patients ([1.36-1.10]/[1.36-1.00] = 0.72 ). After further adjustment for tumor size, neighborhood SES, insurance status, and geographic region, the survival difference between AA and white patients was essentially eliminated (adjusted hazard ratio, 1.03; 95% confidence interval, 0.87-1.21). In contrast, adjustment for undergoing LT was found to have little effect on the results for API patients, who consistently had better survival than white patients. Each 1-cm increase in tumor size was associated with a 6% increase in the risk of death at 5 years. Insurance status was also found to be strongly associated with survival; compared with patients with private insurance, uninsured patients had a 61% increase in their risk of death at 5 years, and patients with Medicaid insurance had a 25% higher risk of death.

Table 4. Multivariate Analysis of Racial Differences in Survival Among LT-Eligible Patients With Hepatocellular Carcinoma, 1998-2002 (n = 2412)
CharacteristicMultivariate HR for Death at 5 Years (95% CI)
Model 1Model 2Model 3
  1. Abbreviations: 95% CI, 95% confidence interval; HR, hazard ratio; LT, liver transplantation.

Race   
 White1.00 (referent)1.00 (referent)1.00 (referent)
 African American1.36 (1.17-1.59)1.10 (0.94-1.28)1.03 (0.87-1.21)
 Hispanic1.08 (0.91-1.27)1.02 (0.86-1.20)0.89 (0.75-1.06)
 Asian/Pacific Islander0.70 (0.59-0.83)0.60 (0.50-0.71)0.59 (0.49-0.71)
 Age (1-y increase)1.03 (1.03-1.04)1.02 (1.01-1.02)1.02 (1.01-1.02)
Sex   
 Male1.00 (referent)1.00 (referent)1.00 (referent)
 Female0.89 (0.80-1.00)0.89 (0.80-1.00)0.88 (0.79-0.99)
 LT   
 No 1.00 (referent)1.00 (referent)
 Yes 0.27 (0.23-0.32)0.29 (0.25-0.34)
 Tumor size (1-cm increase)  1.06 (1.02-1.09)
Percentage of residents without high school degree, %
 <14.0  1.00 (referent)
 14.0−19.9  0.96 (0.82-1.12)
 20.0−28.9  1.09 (0.93-1.28)
 ≥29.0  1.08 (0.89-1.31)
Median household income   
 ≥$46,000  1.00 (referent)
 $35,000−$45,999  1.11 (0.95-1.29)
 $30,000−$34,999  1.07 (0.90-1.27)
 <$30,000  1.13 (0.92-1.38)
Insurance status   
 Private  1.00 (referent)
 Uninsured  1.61 (1.20-2.15)
 Medicaid  1.25 (1.05-1.49)
 Medicare  1.11 (0.97-1.27)
 Other government  1.06 (0.50-2.23)
Geographic region, %   
 Northeast  1.00 (referent)
 Midwest  0.98 (0.84-1.15)
 South  1.05 (0.91-1.21)
 West  1.05 (0.90-1.24)

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. FUNDING SOURCES
  7. REFERENCES

Using national hospital-based cancer registry data from nearly 8000 adults with small (≤ 5 cm), nonmetastatic HCC diagnosed between 1998 and 2007, we found that there were large and persistent racial/ethnic differences in the receipt of LT among patients with HCC. Among HCC patients diagnosed between 2003 and 2007 who were eligible for LT, AA patients with HCC were 36% less likely than white patients to undergo LT, Hispanic patients were 14% less likely, and API patients were 42% less likely. These results changed only slightly after adjustment for age, sex, neighborhood SES, insurance status, geographic region, tumor size, and comorbid medical conditions. Previous studies using national datasets have been limited to patients residing in SEER regions or patients registered on the LT waiting list. Further analyses suggest that racial/ethnic differences in the use of LT among LT-eligible patients with HCC account for racial disparities in survival. This appears to be the first unrestricted, nationwide study of racial/ethnic disparities in the use of LT for HCC.

The findings of the current study are in agreement with previous studies that have found that AA and API patients are substantially less likely than white patients to undergo LT for HCC.7-9 Our finding that AA patients with HCC had poorer survival than their white counterparts is in agreement with the findings of Artinyan et al,10 who also reported that survival differences persisted among patients who underwent LT. In contrast to Moylan et al,12 who reported that racial disparity in the use of LT was decreasing over time, we found no significant change in racial/ethnic disparities with regard to the use of LT for HCC between 1998 and 2007. However, Moylan et al studied patients with all etiologies of end-stage liver disease who were on the LT waiting list.12

The current study has several potential limitations. Because we did not have the information required to calculate each patient's MELD score, we used closely related comorbidity data regarding liver disease, renal failure, coagulopathy, and other conditions, which were strong predictors of undergoing LT, although they were not as accurate as the MELD scores. Although we excluded many patients with missing data regarding stage of disease or tumor size from the current study, the percentage of HCC patients who were excluded for these reasons was nearly identical to that noted in a previous study using SEER registry data.8 The NCDB includes clinical, treatment, and demographic information for approximately 70% of all incident cancer cases in the United States; such data are not centrally available for the state population-based registries. Although data regarding insurance status collected in the NCDB has not been independently validated, it is highly complete and is abstracted from hospital records, in which billing data are likely to be accurately recorded. In addition, there were important patient characteristics that we were unable to consider, such as HCV serostatus,19 HBV antigen positivity,20 α-fetoprotein level,21 diabetes status,22 or dialysis requirement before LT.23 We also had no available data concerning whether individual patients were referred to an LT program. It has been suggested that racial/ethnic differences in referral to a LT center may be one of the most important factors underlying the disparity in access to LT for minority patients.24

The cost of LT may be an important barrier for low-income patients; in 2006, billed costs for LT were estimated to be $449,200 within the first year, and even patients with private insurance may draw from savings accounts or other assets to pay for noncovered expenses.25 AA families generally have lower wealth and assets compared with white families at similar levels of income.26 Studies of surgical treatment of other cancers have consistently found lower rates of surgical treatment among AA patients, even after controlling for clinical and sociodemographic factors; contributing factors may include cultural beliefs as well as a lack of trust in the medical system. In the current study, AA patients had the highest percentage of individuals who did not undergo surgery. Another potential explanation for the lower receipt of LT among API patients may be that more of these individuals underwent surgical resection instead of LT. Among API patients, the etiology of HCC is often chronic HBV infection, and many patients with chronic HBV infection will develop HCC before cirrhosis, which enables surgical resection to be performed, whereas other patients with established cirrhosis may not be candidates for surgical resection. Because of limitations in our comorbidity data, we were not able to test this hypothesis in the current study population; however, we did not observe large racial/ethnic differences in the prevalence of the liver disease comorbidity category.

Despite limitations, the results of the current study provide substantial evidence that there are large and persistent racial/ethnic disparities in the use of LT for HCC in the United States. These disparities were not explained by sociodemographic or clinical factors, although insurance status was found to be an important independent predictor of LT use as well as survival. Although the causes of racial/ethnic disparities in the receipt of LT for HCC are complex and likely to be difficult to change, the current study results suggest that a major benefit of reducing or eliminating these disparities in the receipt of LT may be a substantial reduction in or even the elimination of the racial/ethnic survival disparities observed among patients with HCC.

FUNDING SOURCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. FUNDING SOURCES
  7. REFERENCES

No specific funding was disclosed.

CONFLICT OF INTEREST DISCLOSURES

The authors made no disclosures.

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. FUNDING SOURCES
  7. REFERENCES
  • 1
    Altekruse SF, Kosary CL, Krapcho M, et al, eds. SEER Cancer Statistics Review, 1975-2007. Available at: http://seer.cancer.gov/csr/1975_2007/. Accessed May 15, 2010.
  • 2
    American Cancer Society. Cancer Facts & Figures 2010. Atlanta, GA: American Cancer Society; 2010.
  • 3
    El-Serag HB, Davila JA, Petersen NJ, McGlynn KA. The continuing increase in the incidence of hepatocellular carcinoma in the United States: an update. Ann Intern Med. 2003; 139: 817-823.
  • 4
    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: 693-699.
  • 5
    Figueras J, Jaurrieta E, Valls C, et al. Survival after liver transplantation in cirrhotic patients with and without hepatocellular carcinoma: a comparative study. Hepatology. 1997; 25: 1485-1489.
  • 6
    Ioannou GN, Perkins JD, Carithers RL Jr. Liver transplantation for hepatocellular carcinoma: impact of the MELD allocation system and predictors of survival. Gastroenterology. 2008; 134: 1342-1351.
  • 7
    Yu JC, Neugut AI, Wang S, et al. Racial and insurance disparities in the receipt of transplant among patients with hepatocellular carcinoma. Cancer. 2010; 116: 1801-1809.
  • 8
    Siegel AB, McBride RB, El-Serag HB, et al. Racial disparities in utilization of liver transplantation for hepatocellular carcinoma in the United States, 1998-2002. Am J Gastroenterol. 2008; 103: 120-127.
    Direct Link:
  • 9
    Sonnenday CJ, Dimick JB, Schulick RD, Choti MA. Racial and geographic disparities in the utilization of surgical therapy for hepatocellular carcinoma. J Gastrointest Surg. 2007; 11: 1636-1646.
  • 10
    Artinyan A, Mailey B, Sanchez-Luege N, et al. Race, ethnicity, and socioeconomic status influence the survival of patients with hepatocellular carcinoma in the United States. Cancer. 2010; 116: 1367-1377.
  • 11
    Nguyen GC, Segev DL, Thuluvath PJ. Racial disparities in the management of hospitalized patients with cirrhosis and complications of portal hypertension: a national study. Hepatology. 2007; 45: 1282-1289.
  • 12
    Moylan CA, Brady CW, Johnson JL, Smith AD, Tuttle-Newhall JE, Muir AJ. Disparities in liver transplantation before and after introduction of the MELD score. JAMA. 2008; 300: 2371-2378.
  • 13
    Commission on Cancer. Standards of the Commission on Cancer, Volume II: Registry Operations and Data Standards (ROADS): Revised January 2002. Chicago, IL: American College of Surgeons; 2002.
  • 14
    Commission on Cancer. Facility Oncology Registry Data Standards (FORDS): Revised June 2007. Chicago, IL: American College of Surgeons; 2007.
  • 15
    Fleming ID, Cooper JS, Hensen DE, et al, eds. AJCC Cancer Staging Manual. 5th ed. Philadelphia, PA: Lippincott-Raven; 1997.
  • 16
    Greene FL, Page DL, Fleming ID, et al, eds. AJCC Cancer Staging Manual. 6th ed. New York: Springer-Verlag; 2002.
  • 17
    Fritz A, Percy C, Jack A, et al, eds. International Classification of Diseases for Oncology. 3rd ed. Geneva, Switzerland: World Health Organization; 2000.
  • 18
    Elixhauser A, Steiner C, Harris R, Coffey RM. Comorbidity measures for use with administrative data. Med Care. 1998; 36: 8-27.
  • 19
    O'Leary JG, Randall H, Onaca N, Jennings L, Klintmalm GB, Davis GL. Post-liver transplant survival in hepatitis C patients is improving over time. Liver Transpl. 2009; 15: 360-368.
  • 20
    Jiang L, Yan LN. Current therapeutic strategies for recurrent hepatitis B virus infection after liver transplantation. World J Gastroenterol. 2010; 16: 2468-2475.
  • 21
    Shin WY, Suh KS, Lee HW, et al. Prognostic factors affecting survival after recurrence in adult living donor liver transplantation for hepatocellular carcinoma. Liver Transpl. 2010; 16: 678-684.
  • 22
    Samuelson AL, Lee M, Kamal A, Keeffe EB, Ahmed A. Diabetes mellitus increases the risk of mortality following liver transplantation independent of MELD score. Dig Dis Sci. 2010; 55: 2089-2094.
  • 23
    Dellon ES, Galanko JA, Medapalli RK, Russo MW. Impact of dialysis and older age on survival after liver transplantation. Am J Transplant. 2006; 6: 2183-2190.
  • 24
    Kemmer N, Neff GW. Liver transplantation in the ethnic minority population: challenges and prospects. Dig Dis Sci. 2010; 55: 883-889.
  • 25
    United Network for Organ Sharing. What Every Patient Needs to Know. Available at: http://www.unos.org/SharedContent Documents/WEPNTK2008.pdf. Accessed August 15, 2010.
  • 26
    Gittleman M, Wolff EN. Racial differences in patterns of wealth accumulation. J Hum Resour. 2004; 39: 193-227.