Discussion
HCC represents the most frequent indication for LT for malignancy. In cirrhotic patients, it can be difficult to differentiate liver nodules such as HCC, iCCA or HCC-CC with current imaging tools, and liver biopsy may not be possible in high-risk patients. Moreover, these entities share similar predisposing factors, rendering their differentiation more difficult. Patients with advanced decompensated cirrhosis, with an indication of liver transplant, may also have nonspecific liver nodules considered to be dysplastic or suspicious of HCC not exceeding transplant criteria. These patients are not usually biopsied due to the high risk of complications. For this reason, there are a number of patients who are found to have iCCA or HCC-CC on pathology examination after transplant.
The experience with LT for cirrhotic patients diagnosed with iCCA on pathology examination (incidental tumors or patients misdiagnosed with HCC) or patients transplanted with a known iCCA is limited and available studies included few patients. This fact may explain why, to our knowledge, the present study is one of the few to analyze the risk factors for tumor recurrence post-LT in cirrhotic patients with iCCA on pathology examination and is probably the largest series with these characteristics published to date although the number of patients remains limited. Other studies that attempted to describe risk factors included patients with both iCCA and extrahepatic cholangiocarcinoma and patients with and without liver cirrhosis. Though risk factors for tumor recurrence in patients with iCCA may be expected to be similar to those for HCC, this has not been analyzed to date.
In the current study, we made an exploratory analysis of possible risk factors for tumor recurrence. Due to the small number of patients and variables to be analyzed, we decided to perform only a univariate analysis, the results of which revealed interesting data.
Tumor size on pathology examination had an impact on tumor recurrence after LT. A cutoff of 2 cm was found to be a prognostic factor in patients with iCCA. Differences in tumor size accepted for HCC are probably related to the well-known more aggressive biology of these tumors. TTV also had an impact on tumor recurrence. All these data suggest that, as with HCC, patients with "very early" iCCA have a lower risk of recurrence.
Due to the small number of patients and tumors and the great heterogeneity in preoperative tumor treatments, no conclusions can be drawn as to the effectiveness of these treatments in patients with iCCA.
To date, iCCA has been a formal contraindication for LT at most transplant centers due to poor results in terms of tumor recurrence and patient survival. In our study, we attempted to identify a group of cirrhotic patients that could benefit from LT even if they had an iCCA. Interestingly, despite the 5-year cumulative risk of recurrence for the whole group being high as previously described, patients with tumors ≤2 cm did not present tumor recurrence with a minimum follow-up of approximately 2 years. In fact, although the number of patients was low and four of them had incidental tumors, those with small tumors had a much lower risk of recurrence than patients with larger tumors. This was confirmed in our univariate analysis where tumor size and TTV had an impact on recurrence.
The survival achieved for the entire group cannot be accepted in the current era for patients transplanted for malignancy. However, the survival of patients with "very early" iCCA was significantly encouraging: 73% survival at 5 years. This is an excellent survival rate that compares with that of patients with HCC transplanted within Milan criteria and better than the 5-year survival of patients with HCC transplanted with extended criteria.
Previous reports attempted to demonstrate that patients with iCCA at an early or very early stage may be good candidates for LT. In a recent study by Friman et al including both extrahepatic cholangiocarcinoma and iCCA, it was reported that patients with a TNM stage 1 or 2 had better 5-year survival than those with a stage 3 or 4. Another case report from Sotiropoulos et al suggested that patients with small iCCA could benefit from LT.
Our study offers new information regarding cirrhotic patients with iCCA confirmed on pathology study. Despite the limited number of patients, it provided intriguing information. On the basis of our results, the question of whether patients with single iCCA ≤2 cm should still be excluded from LT could be open to debate. Certainly, if these data are confirmed by other groups or with a larger cohort, there should be a reappraisal of the indication of LT in cirrhotic patients with small iCCA. The most important challenge in that case would be to be able to identify an iCCA in the preoperative setting. In fact, we found four of eight patients with tumors ≤2 cm to be incidentals. Nevertheless, as imaging improves the number of small lesions identified in the preoperative setting increases and studies are starting to identify specific radiologic features of iCCA in cirrhotic patients. We expect, therefore, that the number of incidental tumors will decrease. According to current guidelines, if a nodule in a cirrhotic patient does not have specific characteristics on imaging, the patient should undergo a biopsy and this may force to balance the risks and benefits of liver biopsy in those groups still reluctant to such a diagnostic intervention.
One of the open issues that should be considered is that we cannot affirm that the outcome of patients preoperatively diagnosed of an iCCA will be the same as that of patients diagnosed on pathology examination. Furthermore, no intention-to-treat analysis is feasible and it may be that some patients with undiagnosed iCCA, but wrongly suspected of having HCC, may have progressed rapidly on the WL and been excluded. As a consequence, it would be optimal to have the same analysis carried out by other groups and, if the message is reinforced, conduct a prospective study in the subgroup of patients with iCCA diagnosed at a very early stage.
This study had certain limitations, the most important of which was the number of patients analyzed, particularly taking into account those with tumors ≤2 cm. Although the outcome in those patients was very good, only eight were analyzed and, in fact, half of them were incidentals. Moreover, from a statistical point of view, hazard ratios may be magnified due to the scant number of cases and events, as for patients with tumors ≤2 and >2 cm. This may produce instability of estimations, although it would not affect the direction of the association. Thus, it can be argued that the results of this experience may not impact on practice. Nevertheless, this was the first multicenter attempt to analyze the outcome of these small iCCA after LT and may lead to future prospective analyses that could really impact on patients diagnosed with such tumors preoperatively who currently should not access transplant lists.
In summary, the present study was the first attempt to ascertain the risk factors for tumor recurrence in cirrhotic patients found to have an iCCA on pathology examination. The risk factors for tumor recurrence are similar to those described for HCC: tumor size, microvascular invasion and tumor differentiation. Cirrhotic patients with iCCA ≤2 cm achieved excellent 5-year survival, and validation of these findings by other groups may address the current exclusion of such patients from transplant programs.