Liver Transplantation and Incidentally Found Hepatocellular Carcinoma in Liver Explants: Need for a New Definition?

Abstract

Background. 

“Incidentally” identified hepatocellular carcinoma (iHCC) in liver explants after liver transplantation (LTx) is a frequently reported finding, which is characterized with a good prognosis. The purpose of this study was to evaluate the outcome of patients with these tumors in our series and in literature reports, and to compare their prognosis to that of HCC diagnosed preoperatively.

Methods. 

From April 1998 to December 2003, 432 patients underwent deceased-donor LTx at our center for nonmalignant indications. An additional 31 patients with a preoperative known HCC (pkHCC) received deceased-donor grafts. A literature search was performed intending to estimate the incidence of iHCC in liver explants and the outcome after LTx.

Results. 

iHCC was found in 5 of the 432 patients. All five patients are currently alive without evidence of tumor recurrence after a median follow-up of 43 months. On the other hand, in the group of the 31 patients with pkHCC, 22 of them are at the moment alive in a median follow-up of 28 months. When comparing the two groups, no difference in survival could be found (P=0.1419 in log-rank test). Literature reports of 705 instances with iHCC over the past 20 years showed a statistical “better survival” in only 24 cases.

Conclusion. 

Literature reports showed a remarkable “deviation” of the expected tumor characteristics for the iHCC. Obviously, this is because of a widely characterization of iHCC, including also tumors which are rather undetected HCC during the waiting time to LTx. A more precise definition for the iHCC is needed.

Liver transplantation (LTx) is the only treatment for end-stage cirrhosis (^1–2). Despite improvements in imaging techniques, the diagnosis of hepatocellular carcinoma (HCC) during the pretransplant evaluation remains difficult, resulting in a considerable rate of undetected HCC in liver explants. Unsuspected HCC found at the time of histological examination of liver explants of patients transplanted for benign diseases is defined as “incidental HCC” (³). Although there is no current worldwide consensus as to the “optimal” listing criteria for patients with known HCC and cirrhosis, some have suggested that undetected “incidentally found” HCC is associated with a better prognosis (^4–5), similar to that of patients transplanted for nonmalignant diseases.

We herein describe our experience with liver transplantation and incidental HCC (iHCC) and we present the findings of a literature search about LTx and incidentally found HCC.

PATIENTS AND METHODS

From April 1998 to December 2003, 463 patients underwent deceased donor LTx at our center. Thirty-six of these recipients had HCC confirmed at the time of pathologic examination.

Five of them (14%), transplanted with no preoperative diagnosis of malignancy, had HCC detected “incidentally” in the liver explants (group 1). All patients had a diagnosis of liver cirrhosis and were evaluated for liver transplantation according to our standard evaluation protocol. The evaluation procedure included computed tomography of the liver (2-mm slices), abdominal ultrasonography, and estimation of the tumor markers alpha fetoprotein (AFP), carbohydrate antigen 19-9, and carcinoembryonic antigen prior to listing. During the waiting period, abdominal ultrasonography was repeated at least every 3 months. Because tumor markers were obtained only once during the evaluation process in cases where no tumor was suspected, serial values were not available for those patients with incidental HCC. By suspicion of HCC during the waiting time to LTx, patients were further evaluated with abdominal and thorax computed tomography, bone scintigraphy and re-evaluation of AFP-levels. Patients with a suspicion of HCC during the waiting time to LTx were not included in this group.

An additional 31 patients with a preoperative diagnosis of HCC received deceased-donor organs (group 2). In this patients’ group, AFP was estimated every 3 months during the waiting time to LTx. Abdominal computed tomography was performed at the time of evaluation to LTx, whereas abdominal ultrasonography was routinely repeated at least every 3 months. If there was a suspicion for further tumor growth during the waiting time, abdominal and thorax computed tomography, as well as bone scintigraphy, were performed. Ten patients underwent transarterial chemoembolization as bridging treatment prior to LTx, whereas 10 additional patients underwent radiofrequency ablation. According to the radiological imaging, all patients were meeting the Milan criteria at the time of listing and during the waiting time to LTx.

All patients received a deceased donor full-size LTx. Careful pathologic study (1-cm slices) of all livers removed was performed by the same pathologist. Postoperatively standard cyclosporine-based immunosuppression was employed in all cases. In every instance, pre- and postoperative care was rendered at our outpatient liver transplant clinic. Tumor staging for HCC was performed every 3 months during the first year, every 6 months during the second year, and yearly thereafter. No tumor-related therapy was administered postoperatively.

The outcome of the patients with incidental tumors (group 1) was compared to that of the 31 patients with clinically detectable HCC prior to LTx (group 2).

Search Strategy and Selection Criteria

On March 2005, PubMed and Medline were accessed and searched to estimate the incidence and the outcome of iHCC in liver explants after LTx. “Hepatocellular carcinoma,” “liver transplantation,” “incidentally found,” and “liver explants” were entered as keywords both independently and in multiple combinations. Only English-language papers published between 1985 and 2005 were considered. In instances of multiple studies from a single institution, the most recent or more informative publication was chosen.

Statistical Analysis

Survival was calculated using the Kaplan-Meier method, with statistical significance determined by the log rank test based on the exact permutation distribution. The Cochran-Mantel-Haenszel-Test was used to compare the two groups based on the data found in the literature. For the combination of P values, the truncated product method (⁶) was used with tau=0.05. Thus, studies that report a nonsignificant result without giving the actual P value could be included in the meta-analysis. P values ≤0.05 were considered significant. The software SAS (SAS Institute, Cary, NC) and StatXact (Cytel, Cambridge, MA) were used to perform the statistical analysis.

RESULTS

HCC was incidentally found in four men and one woman in group 1. Three patients had been transplanted for cirrhosis associated with viral hepatitis (two for hepatitis C and one for hepatitis B), and two for alcoholic cirrhosis. Median age was 52 years (range 43–61 years) and median waiting time to LTx was 316 days (range 45–358 days). AFP values prior to LTx were normal in all cases. The last computed tomography study was performed a median of 90 days prior to LTx. HCC was not suspected at any time during the evaluation process. All patients were Child-PughTurcotte class B at the time of transplantation. Histology showed solitary HCC in three cases (tumor diameters of 1.4 cm, 1.8 cm and 2.5 cm respectively) and multifocal bilobar tumors <1 cm in diameter in the remaining two cases. None had vascular invasion detected. HCC was moderately differentiated in four cases and well differentiated in one case. Three cases were stage I and two stage II according to the classification criteria of the Union Internationale Contre le Cancer. Based on the pathological findings, the three solitary tumors met the Milan criteria whereas the two multifocal ones exceeded them. All five patients experienced uneventful posttransplant courses and were discharged a median of 21 days after transplantation (range 18–25 days). All patients are alive 83, 73, 43, 23, and 23 months after LTx, with a median follow-up of 43 months. No patient developed tumor recurrence.

In the second group of the 31 patients with a preoperative diagnosis of HCC, mean waiting time to LTx was 154±185 days. There were 22 men and 9 women with a mean age of 54.5±11.69 years. Cirrhosis was documented in all these patients, and was attributable to hepatitis C infection, hepatitis B infection, combination of hepatitis B and C infection and alcohol abuse in 42% (13/31 patients), 32% (10/31 patients), 6% (2/31) and 10% (3/31 patients) of such cases, respectively. The remaining 10% had cirrhosis of unknown etiology. Pathological evaluation of the explanted livers showed tumors meeting the Milan criteria in only 39% (12/31) of the instances. In 19 patients pathology revealed tumors exceeding the Milan criteria, showing a corresponding underestimation of these tumors in the radiological evaluation. Twenty-two patients are at the moment alive. The overall survival rates at 1 and 3 years were 91% and 78%, respectively (Fig. 1) with a median follow up of 28 months (range 13–77 months).

FIGURE 1.: Comparison of outcomes between patients with iHCC and patients with previously known HCC (pkHCC); P=0.1419 in log-rank test.

The overall survival of patients with iHCC showed no statistically significant difference when compared to that of patients with known HCC (pkHCC) prior to transplantation (P=0.1419 in log-rank test; Figure 1).

Review of the Literature with Reference to Incidentally Found HCC in Liver Explants

Although there are many reports in the literature on outcome after liver transplantation for HCC, there is limited information on iHCC. Thirty-one studies (ours included) fulfilling the criteria described in Patients and Methods were identified (^{3, 5, 7–34}) (Table 1). The incidence of such tumors among transplanted patients with no known HCC ranged from 1% to 17.5% in liver explants. Incidental HCC 3–5 cm in mean diameter was observed in a total of 25 patients by Yao et al. and Leung et al. (^{17, 30}). Tumor recurrence was observed in 146 patients out of 382 patients with iHCC with adequate date (Table 1). Survival of patients with iHCC was documented to be significantly better than that of patients with pkHCC in only 24 (3%) of the 705 patients with iHCC reviewed in the present study (^{19, 24}). On the other hand, the outcome of patients with iHCC was not found to be significantly better in 343 (49%) patients. In the remaining 338 (48%) patients of Table 1, no comparison of outcome was reported.

TABLE 1: Review of the published literature addressing incidentally found HCC in liver explants after liver transplantation

The incidence of recurrences was higher in the group of pkHCCs and this difference was significant in CochranMantel-Haenszel-Test (P<0.0001 between 14 studies with adequate information). Statistical analysis showed as well a significant better survival in the group of iHCCs (P<0.05 using the truncated product method between the 12 studies with corresponding data in the present literature review).

DISCUSSION

The detection of small HCC in the setting of end-stage liver cirrhosis remains a diagnostic challenge for physicians. It is many times almost impossible to distinguish small HCC from macroregenerative nodules, frequent and benign lesions in cirrhotic livers, and from dysplastic nodules, presently considered to be premalignant lesions (³). Krinsky et al. showed an overall sensitivity of 33% for magnetic resonance imaging, with only 52% of lesions between 1 and 2 cm, and 4% of lesions under 1 cm correctly classified (³⁵). Lopez Hanninen et al., evaluating computed tomography, reported a sensitivity of 20% for tumors between 1 and 2 cm (³⁶). Similar findings were reported by Bhartia et al., who concluded that identification of tumors smaller than 1 cm is still limited (³⁷). As already reported, sensitivity of radiological imaging was proved especially poor for tumors between 1–2 cm and <1 cm (21% and 0% respectively) in a retrospective study of 70 patients with HCC in cirrhosis undergoing liver transplantation in our center (³⁸). Based on these observations, it is still not possible to adequately and safely screen for HCC smaller than 2 cm.

The literature over the past 20 years is not completely consistent as to whether these “incidentally found” tumors have indeed a better prognosis than HCC detected preoperatively (^{3, 5, 7–34}) (Table 1). Data are inhomogeneous, contradictory, and incomplete. We evaluated 31 series, for a total of 705 patients with reported iHCC. Although iHCC is expected to be a tumor with characteristics meeting the Milan criteria (³⁹), the literature review proved a remarkable deviation concerning these tumor characteristics: 1) multifocal iHCC was frequently described (in 54% of the cases in the report of Cillo et al., in 48.3% in the report of Klintmalm et al., in 33% in the report of Margarit et al., in 23% in the report of González-Uriarte et al., in 15.2% in the report of Khakhar et al.); and 2) iHCC with vascular invasion was documented in a median rate of 15% (in 31% of the cases in the report of Achkar et al., in 30% in the report of Yao et al., in 30% in the report of Khakhar et al., in 16% in the report of Klintmalm et al., in 13% in the report of Margarit et al.). The report of Klintmalm in his prospective, randomized multicenter trial showing iHCC in 40% of the 169 patients analyzed (¹³) indicates obviously a “problem of definition” or a problem of inadequate evaluation of the patients during the waiting time to LTx.

Statistical analysis of the literature data of Table 1 showed a significant difference for the recurrences between the 14 studies with corresponding data (^{3, 5, 7, 10, 11, 13, 14, 17, 19–21, 25, 29}). The incidence of recurrences was higher in the group of pkHCC (P<0.0001 in Cochran-MantelHaenszel-Test). Statistical analysis showed as well a significant difference for the survival between the 12 studies with corresponding data (^{10, 13–14, 17, 19, 23–26, 29, 30}). Survival was found better in the group of iHCC, when combining the 12 available P values using the truncated product method (P<0.05 in truncated product method). Although the “favorable outcome” of patients with iHCC in comparison to pkHCC is not well documented in the literature with an observed better outcome of only 24/705 patients reported, statistical analysis of the data of the literature review showed significantly better results for patients with iHCC concerning tumor recurrence and survival.

Although the literature about iHCC provides incomplete and only sporadic data, there is a remarkable “deviation” of the expected tumor characteristics for the iHCC. Obviously, this is because of a widely characterization of iHCC between LTx centers as such, including also tumors which are rather undetected HCC during the waiting time to LTx. Because HCC was not suspected prior to LTx in these cases of LTx for “benign disease,” neither tumor-specific nor intensive evaluations during the waiting period prior to LTx were routinely undertaken. Computed tomography or magnetic resonance imaging studies and tumor markers were obviously not routinely obtained except in cases of known malignancy.

Hepatocellular carcinomas will continue to be found incidentally after LTx for “benign disease.” This trend will persist as long as imaging techniques are not able to detect small tumors in cirrhotic livers, and as long as patients with cirrhosis and no known malignancy are not evaluated and restaged as intensively as tumor patients. Expansion of the tumor staging for all patients with cirrhosis waiting for LTx seems relatively impracticable. Given the above mentioned “problem of definition,” many of the reported in the literature “incidental” HCCs are rather “undetected” HCCs because of the long waiting time and the absence of intensive tumor staging. As a consequence of this situation, the self-evident “better survival” of patients with iHCC is not unequivocally documented in the literature.

According to the results of our review, the widely accepted definition of iHCC as “unsuspected HCC found at the time of histological examination of liver explants of patients transplanted for benign diseases” seems insufficient to distinguish between iHCC and undetected HCC. This definition could result to misinterpretation that patients with iHCC and pkHCC share same outcomes, situation which is not in accordance with the clinical experience. A consensus determining a concrete art of objective radiological evaluation (computed tomography or magnetic resonance imaging) and an acceptable intervening period of time between last radiological evaluation and LTx could guide to a better separation between iHCC and undetected HCC and a better assessment of the patients’ survival.