The development of de novo hepatocellular carcinoma in patients on a liver transplant list: Frequency, size, and assessment of current screening methods

Authors

  • David H. Van Thiel,

    Corresponding author
    1. Department of Medicine (Section of Hepatology), Loyola University Medical Center, Stritch School of Medicine, Loyola University, Chicago, IL
    2. Department of Surgery (Solid Organ Transplantation), Loyola University Medical Center, Stritch School of Medicine, Loyola University, Chicago, IL
    • Hepatology and Liver Transplantation, Physicians Office Building, Suite 525, St. Luke's Medical Center, 2801 W. Kinnickinnic River Parkway Milwaukee, WI 53215
    Search for more papers by this author
    • Telephone: 414-385-1916

  • Sherri Yong,

    1. Department of Pathology, Loyola University Medical Center, Stritch School of Medicine, Loyola University, Chicago, IL
    Search for more papers by this author
  • S. David Li,

    1. Department of Medicine (Section of Hepatology), Loyola University Medical Center, Stritch School of Medicine, Loyola University, Chicago, IL
    Search for more papers by this author
  • Marc Kennedy,

    1. Department of Medicine (Section of Hepatology), Loyola University Medical Center, Stritch School of Medicine, Loyola University, Chicago, IL
    Search for more papers by this author
  • John Brems

    1. Department of Surgery (Solid Organ Transplantation), Loyola University Medical Center, Stritch School of Medicine, Loyola University, Chicago, IL
    Search for more papers by this author

Abstract

Chronic end stage liver disease is the most frequent indication for liver transplantation. Individuals with end stage cirrhosis, and therefore individuals on liver transplant lists, are at increased risk of developing a hepatic cancer. Those individuals on liver transplant lists also may represent the best group available for evaluating the current methods for screening and surveillance for the development of hepatic cancer as an examination of the explant liver provides a gold standard for tumor assessment. Assuming that only tumor free individuals were screened at the onset of this study, the data obtained enables one to determine the frequency of new hepatic cancers since listing and evaluate the positive and negative predictive values of each assessment method over the surveillance interval. All patients listed for liver transplantation with end stage chronic liver disease, who did not have a hepatoma at the time of transplant listing, were followed and assessed for the development of a hepatic cancer while on the waiting list. The screening techniques utilized included quarterly α fetoprotein (αFP) determinations and ultrasound (US) studies as well as semi-annual triple phase computed tomography (CT) scans of the liver. αFP failed to identify any cases of de novo hepatic cancer in patients waiting for a liver transplant. In contrast, US and especially CT scanning with intravenous contrast identified new hepatic masses. The later method, which identified early enhancing mass lesions, was the more valuable method at identifying masses that subsequently were shown by pathologic examination of the explant liver to be hepatic cancers. However, only 14 of 20 individuals found to have a de novo tumor were identified by this method. Once identified however, the treatments utilized for hepatic tumor ablation while waiting for a transplant appear to be effective with a mean of 57.8±8.3% necrosis of the treated masses being identified at the time of explant examination. In conclusion these data suggest that:

  • 1The development of a hepatocellular carcinoma (HCC) in an individual on a transplant list is not rare and occurs in as many as 20% of cases;
  • 2The most effective method for the detection of de novo HCC appears to be semi-annual triphasic CT scan with the identification of a new early enhancing lesion; and
  • 3Once recognized, the presence of the tumor enables the individual to move up on the waiting list as result of the additional model endstage liver disease (MELD) points allowed for individuals with HCCs. (Liver Transpl 2004;10:631–637.)

The major etiologic agents identified for the development of hepatocellular carcinoma (HCC) in the western world are chronic hepatitis C virus (HCV) infection and chronic alcohol abuse.1–16 Worldwide, particularly in Asia and Africa, the leading etiologic agent for the development of HCC is the presence of a chronic hepatitis B infection.3, 4, 7, 8, 12, 13 Chronic hepatitis C appears to be responsible for the recent increase in prevalence of HCC in the United States5, 6 and in other unique groups.9–11 HCC is also a well-recognized complication of cirrhosis in individuals with the various toxin exposures particularly alcohol but also other environmental toxins.1, 2, 9, 16, 17 Finally, HCC occurs in cases of advanced endstage liver disease regardless of the etiology.1, 2, 9, 10, 14, 15 Thus, candidates for liver transplantation on a waiting list are a unique population at risk for the development of HCC. Despite this fact, the rate of de novo HCC development in liver transplant candidates listed for transplantation is currently unknown. Moreover, the efficacy of current screening and surveillance modalities for the detection of an HCC is such a high-risk population for de novo HCC development are also unknown.

Abbreviations:

αFP, α´fetoprotein; US, ultrasound; CT, computed tomography; HCC, hepatocellular carcinoma; MELD, model endstage liver disease; HCV, hepatitis C virus; UNOS, United Network for Organ Sharing; INR, international normalized ratio; HBV, hepatitis B virus; NASH, nonalcoholic steatohepatitis.

Materials and Methods

Subjects

Individuals with endstage liver disease due to any cause, who were evaluated and found to be free of an identifiable HCC and who met United Network for Organ Sharing (UNOS) criteria for listing for liver transplantation, were included in the present study.

Specifically, from October 1998 through July 2003, a total of 300 individuals were evaluated and presented to the liver transplant review board at Loyola University Medical Center. Of these, 282 were listed for transplantation. Fifteen of these cases were identified as having an HCC at the time of listing and five were listed because of fulminant hepatic failure. These cases were eliminated from the subsequent analysis leaving a total of 262 listed liver transplant candidates. Of these, 105 (41%) were transplanted with four individuals receiving two and one individual receiving three transplants. These later cases receiving multiple transplants were eliminated leaving 100 cases for analysis (Figure 1).

Procedures

Evaluation Procedures

The liver transplant evaluation procedures consist of a complete virologic, serologic, and biochemical evaluation for the recognized causes of end stage liver disease. In addition, imaging procedures consisting of a CT of the head, triphase CT of the abdomen, and an US examination of the liver and its vessels as well as the biliary tree are obtained. Finally, αFP, CT, and US studies are obtained to screen for the presence of hepatic cancer.

Surveillance Procedures Utilized for Transplant Listed Cases

Once the individual is listed for transplantation, the routine measures required by UNOS for Mayo Endstage Liver Disease (MELD) scoring (prothrombin time/internation normalized ratio [INR], serum total bilirubin, and serum creatinine)19 are monitored every 2–3 months or more often depending upon the most recent MELD score and UNOS rules.

In addition to these monitoring measures, each listed candidate undergoes continuous surveillance for the presence of hepatic cancer utilizing a quarterly determination of the serum α FP level and an abdominal US examination and a semi-annual triphasic CT scan of the abdomen. Any elevation of the αFP ≥ 200 ng/mL was considered to be highly suggestive of the presence of a primary hepatic cancer. Moreover, a doubling of the initial screening αFP level during surveillance to a value ranging between 50–200 ng/mL was considered to be highly suggestive of the development of an HCC.

The US criteria used to identify a new hepatic lesion consisted of the finding of either a hypoechoic lesion >1 cm in diameter, or a target lesion consisting of a hypoechoic lesion >1.5 cm in diameter with a central hyperechoic area or a mass adjacent to a thrombosed intrahepatic portal vein radicle.

The criteria used to identify a hepatic tumor by CT consisted of the identification of a focal lesion ≥6 mm in diameter with early arterial enhancement.

All of the US and CT scans obtained during the initial screening and subsequent surveillance studies were read by staff radiologists with expertise in abdominal US and CT diagnostic techniques and were faculty of the radiology department of Loyola University Medical Center, Maywood, IL.

Pathologic Examination

The explant liver and hepatic vessels were examined grossly for the presence of any tumor. The most recent CT findings were utilized by the pathologist to identify lesions that were identified prior the transplantation for the presence of tumor. In addition, any lesion recognized by the pathologist but not recognized by the CT or US surveillance studies were examined histologically. Two areas of each identified lesion and two additional randomly chosen areas from each hepatic lobe (16 blocks) as well as three separate randomly selected areas of the intrahepatic portal and hepatic veins were examined histopathologically.

Statistical Analysis

All of the data presented are mean values ≥SEM. The sensitivity and specificity of each surveillance technique (αFP, abdominal US, and the triphasic CT scanning) were calculated at the time of transplantation utilizing the pathologist's identification of a tumor as the gold standard.

Chi square and Kruskal-Wallis tests were used to compare categorical data. A P value less than 0.05 was considered to be significant.

Figure 1.

Distribution of case material in the population studied.

Results

None of the 100 cases followed as part of this investigation were known to have a hepatic cancer at the time of listing. The primary liver disease responsible for the endstage liver disease in those who did not and who did develop an HCC while waiting for a transplant are shown in Table 1. The mean MELD scores as well as the age and gender of the two groups are shown in the Table 2. Fourteen of these 100 (14%) were discovered to have developed a de novo hepatic cancer while on the waiting list as a result of the imaging surveillance procedures utilized. The αFP levels varied over time but both increases and reductions were seen. None of the tumors identified in this study were detected as a result of αFP testing at quarterly intervals. Specifically, the criteria utilized to identify a HCC using αFP failed to identify a single tumor. The 14 cases that had a tumor recognized and treated prior to transplantation had a total of 26 tumors. Most of the additional tumors were small <5 mm satellite lesions that were not seen using the surveillance methods utilized. Those that were treated prior to transplantation demonstrated an average of 57.8 ±8.3% necrosis when the explant liver was examined pathologically (range 0–100%). Only one patient, who appeared to have developed tumor while on the waiting list experienced recurrent tumor. This single patient ultimately died of a rapidly growing highly malignant HCC that prior to transplant was thought to consist of two isolated tumor nodules at opposite poles of the liver (far right lobe segment 6) and in the far left lobe (segment 3) but was found on examination of the explant liver to be a massive tumor involving the entire liver with 80–85% of the hepatic mass consisting of neoplastic tissue.

Table 1. Demographic Characteristics of the Subjects Segregated as to the Presence or Absence of Tumor in the Explant Liver
 No HCC n = 82HCC Identified While on List n = 14HCC Found only after OLTx n = 6
  1. Abbreviations: HCC, hepatocellular carcinoma; HCV, hepatitis C virus; HBV, hepatitis B virus; MELD, model endstage liver disease; NASH, nonalcoholic steatohepatitis.

  2. †Glycogen storage disease, candida cholangitis, α1 anti-trypsin deficiency (ZZ phenotype), and a chronic Budd Chiari syndrome.

  3. Values expressed as mean ± SEM.

Age (yr)49.1 ± 1.260.2 ± 2.657.4 ± 3.1
Gender (M/F)52/3011/35/1
Disease etiology   
 Hepatitis C virus (HCV)2873
 Alcohol induced cirrhosis2622
 Primary biliary cirrhosis700
 Autoimmune hepatitis & cirrhosis400
 Primary sclerosing cholangitis400
 Cryptogenic cirrhosis421
 Hepatitis B virus (HBV)020
 NASH* associated cirrhosis210
 Other misc 4 diseases400
MELD score at listing22.2 ± 2.119.4 ± 3.120.8 ± 3.0
MELD score at time of OLTx30.4 ± 1.726.0 ± 2.922.8 ± 2.0
Time on the waiting list241.1 ± 64.3247.2 ± 53.8146.4 ± 58.1
Table 2. Time on the Waiting List for Those Who Did and Did Not Developed a De Novo HCC
GroupWaiting Time (days)
  1. Abbreviation: HCC, hepatocellular carcinoma.

  2. Values expressed as mean ± SEM (range).

82 cases that did not develop HCC241.1 ± 64.3; (3– 1279)
14 cases that developed HCC after listing but prior to OLTx243.2 ± 21.9; (1– 757)
6 cases identified as having HCC only after OLTx146.4 ± 58.1; (2– 468)

The mean initial MELD scores at the time of listing and at the time of recognition utilizing US and CT to identify a hepatic tumor in these 14 cases were 19.4±3.1 and 21.3±2.6. The time from recognition of the hepatic tumor to the date of liver transplant was 247.2±53.8 days and their immediate mean pretransplant MELD score with the additional points accrued for the presence of a hepatic cancer was 26.0±2.9.

Six additional individuals (6%) were found to have developed a hepatic cancer while on the waiting list prior to transplantation but escaped detection as a result of the surveillance techniques utilized and were identified only after the explant liver was examined pathologically. The initial transplant MELD score for these six individuals was 20.8±3.0; their MELD score at the time of transplantation was 22.8±2.0. Their transplant waiting time was 146.4±58.1 days.

Thus, 20% of those receiving first transplants for end stage liver disease in this series and during this period appeared to develop a de novo hepatic cancer while waiting to be transplanted. The initial and pretransplant MELD scores for these 20 cases were 19.9±1.1 and 24.1±1.1 respectively with a waiting time of 227.1±64 days.

An additional five cases (5%) were thought to have developed an HCC while on the waiting list utilizing the US findings and were treated with ethanol injections but were found to have no evidence of HCC when the explanted liver was examined pathologically. The initial and pretransplant MELD scores for these five cases were 19.8±2.1 and 24.8±2.1 respectively. Their total waiting time on the transplant list was 192.6±59.3 days with 112.2±38.4 of these days on the list after the identification of the putative hepatic cancer.

The initial and immediate pretransplant MELD scores for those 80 cases not developing a hepatic cancer while on the waiting list were 22.2±2.1 and 30.4±1.7 with a mean waiting time of 241.1±64.3 days.

The mean size of the tumors not identified prior to an examination of the explant liver was 3.1±0.8 cm in greatest diameter with a range of 0.1–7.0 cm. In contrast, the tumors that were recognized by the surveillance procedures averaged 2.5±0.4 cm with a range of 0.3–3.5 cm. Each of these tumors were identified by the triple phase CT scanning of the liver followed by lipoidal arteriography and were documented histopathologically after a liver biopsy was obtained from the lipoidal positive area. As noted above, serial αFP levels utilizing the criteria described failed to identify any of these de novo tumors. Abdominal US identified 12 of the 14 tumors recognized by CT and the 5 lesions thought to be a HCC but which were not documented by examination of the explant liver.34, 35 Once identified, the tumors were treated with either ethanol injections or chemoembolization on a regular schedule of monthly treatments followed by a predetermined 3-month post-treatment observation period before consideration of any additional retreatment. Transplantation was not delayed in any patient because of this treatment plan. Thus, if a donor organ became available, the patient was transplanted regardless of how far into the treatment protocol they had progressed. Only four patients completed three scheduled chemoembolization treatments before being transplanted while two were transplanted while receiving treatment that was interrupted because of the availability of a donor organ. Nine received no tumor therapy before their transplant because of medical factors that prohibited treatment such as advanced thrombocytopenia, disease severity, (coagulopathy and/or azotemia), or location of the tumor (Table 3).

Table 3. Time after Identification and Ablative Therapies Utilized in Those Identified as Having a HCC While Awaiting OLTx
Therapeutic ProcedurenMean Applications of the Identified Therapy/Case
  1. n = 14 278.1 ± 69.69; range 2 to 744

  2. Therapies provided to the 17 cases that developed a de novo HCC (proven n = 14 and unproven n = 3) and were treated for the HCC prior to OLTx.

Alcohol injection41
Chemoembolization32.2
Radio frequency ablation11
No therapy9not applicable

The 14 cases de novo tumors detected in these series were all detected by triphase CT as a mass with early arterial enhancement.32, 33, 37 The six cases that had a mass but were not recognized as tumors failed to demonstrate early arterial enhancement and were thought to be hyperplastic nodules. The five cases that were treated with ethanol injections and subsequently found not to be HCCs were all identified by US as hypoechoic masses.34, 35 As noted, the explant liver in these five cases failed to demonstrate any residual tumor. Thus, these five cases represent either tumor areas that were completely eliminated or represented false positive US findings. Assuming the former, the sensitivity and specificity of US were 60% and 93.8% with a positive predictive and negative predictive value of 70.6% and 90.5%. The sensitivity and specificity values for CT scanning for the detection of a new HCC were 70% and 100% with a positive predictive value of 100% and a negative predictive value of 89.9%.

Discussion

Advanced endstage liver disease is the principal indication for liver transplantation in the world.20 Worldwide the three most frequent disease indications for liver transplantation are chronic hepatitis B (HBV) particularly in the Middle East, Asia, and Southern Africa, and chronic hepatitis C (HCV) and alcoholic liver disease in Western Europe and the United States.20–23 HCC is a frequent late complication of advanced cirrhotic liver disease.1, 4, 5, 7, 10, 12, 13, 16 The frequency of HCC in the United States is increasing and this increase appears to be a consequence of the current epidemic of chronic hepatitis C.4, 5 As a result, individuals on liver transplant waiting lists represent a group of patients at very high risk for the development of a HCC and as such would appear to be an ideal group for the evaluation and application of screening and surveillance techniques for the detection of HCC as the results of these procedures can be validated with a pathologic examination of the explant liver.

Moreover, with the recent changes in the UNOS listing criteria for liver transplantation in the United States, individuals with small hepatic cancers such as those likely to be detected as a consequence of surveillance techniques are awarded additional points moving them up on the transplant list ahead of their peers with equivalently advanced liver disease but without an HCC.19–26 This change in UNOS listing criteria has increased the transplant rate for individuals with HCC 3.5 fold and has encouraged an interest in surveillance of individuals on liver transplant waiting lists for HCC at many centers.39–44

In the present investigation, the apparent de novo development of a HCC in individuals on a liver transplant list occurred in 20% of the patients on the waiting list. Of these 20 cases, 14 (70%) were discovered as a result of the surveillance procedures utilized.35, 36 Importantly, the serial CT scans were found to be most useful in this series for the detection of a HCC.32, 33, 37 An additional six (30%) were found to have a HCC when the explant liver was examined pathologically but were not identified by any of these surveillance techniques utilized. None of the pretransplant detected cases were identified as a result of the serial αFP testing utilizing the criteria for tumor detection described in the methods. Twelve of the 14 cases identified as developing a de novo HCC were identified by US as well as CT. Five additional cases were identified as a result of the US surveillance technique and were treated with ethanol injections. None of these later five cases demonstrated residual tumor on pathologic examination of the explant liver and either represent tumor cures or false positive US results. The criteria utilized to identify a HCC using ultrasound were either the detection of a new focal hypoechoic lesion or the finding of thrombosis of an intrahepatic portal vein radical adjacent to a hepatic nodule regardless of its echogenic appearance.

All of the tumors that were identified pretransplantation (70%) were detected as a new early enhancing focal lesions within the liver as part of a triple phase CT scan of the abdomen.32, 33, 37 Six cases (30%) however were not detected using this technique.

Other methods for screening for the development of HCC exist. These include measurement of descarboxyprothrombin, NMR scanning, and PET scanning.38 Only NMR with superparamagnetic iron oxide enhancement has been shown to be more useful than the instruments used in this study.27 This method however is less available and considerably more costly than the procedures utilized in the present investigation and is therefore unlikely to be accepted as a screening or surveillance procedure.

Of interest was the finding that the tumors ultimately identified in these 20 patients were all <5 cm in diameter with one exception and were contained within the liver. Thus, each was expected to be cured as a consequence of their transplant procedure. The single case of diffuse HCC misidentified as two nodules at opposite poles of this liver developed recurrent HCC post transplant and actually had a single massive tumor involving >85% of the liver at the time of explant examination.

The frequency of the surveillance techniques used for identifying a tumor in the present study are consistent with the data available on the growth rates of HCC contained in the literature.28–31 Specifically, the doubling time for HCC has been reported to range from 34.8–496.4 days with a mean of 93.5 days.31

There were important pitfalls in this study. These include the relative small number of cases followed (n=100) and the limited length of follow-up averaging 2/3 of a year. The important positive characteristics of this study however were: 1) the clear definitions of what findings were considered positive for a tumor using each surveillance method; 2) the blinded radiologic readings of the US and CT images, and 3) the use of the imaging studies by the radiologist to identify all tumors present in the explant livers.

In conclusion these data suggest that:

  • 1The development of a HCC in an individual on a transplant list is not rare and occurs in as many as 20% of cases;
  • 2The most effective method for the detection of de novo HCC appears to be semi-annual triphasic CT scan with the identification of a new early enhancing lesion;
  • 3Early recognition and subsequent treatment followed by liver transplantation enables the great majority of these cases to be cured both of their end stage liver disease and their tumor;
  • 4The application of therapeutic procedures directed at tumor ablation in these cases appears to be effective at controlling the growth of the tumors until transplantation can be accomplished.

Ancillary