Please use this identifier to cite or link to this item: http://hdl.handle.net/10362/31474
Title: Decision making in liver transplantation for hepatocellular carcinoma: a study on clinical and molecular prognostic factors
Other Titles: Decisão clínica no transplante hepático por hepatocarcinoma: um estudo de factores de prognóstico clínicos e moleculares
Author: Marques, Hugo Silva Carvalho Pinto
Advisor: Fragata, José
Barroso, Eduardo
Leal, José Pereira
Keywords: Liver transplantation
Hepatocellular carcinoma
Clinical and molecular prognostic factors
Defense Date: 26-Feb-2018
Abstract: ABSTRACT: HCC (hepatocellular carcinoma) is a highly prevalent disease and the second leading cause of cancer-related mortality in the world. Liver transplantation (LT) is currently the best treatment for HCC in cirrhotic patients, even in the case of advanced liver-only disease; nonetheless it carries the problem of organ availability and competition from patients more capable of benefiting from this kind of therapy. New approaches to deal with organ shortage need to be explored. On the other hand, proper selection of patients is imperative. Current clinical morphological models, namely the Milan Criteria, may exclude excellent candidates. Furthermore, patients with early HCC and an aggressive biological behavior are not excluded by current criteria. It is imperative that better selection models are developed, which are more sensitive and specific. Molecular markers can predict the behavior and aggressiveness of HCC. Consequently, it is likely that integrating clinical and molecular features will be the basis of improved decision criteria, achieving a better patient selection for LT due to HCC in the setting of cirrhosis. The main objective of this work was to identify, in cirrhotic patients with HCC, the best candidates for liver transplantation. Specifically, we analyzed the impact of domino liver transplantation in this setting as a means to ameliorate the problem of organ availability. We refined clinical criteria for selection of cirrhotic HCC patients for LT. We further evaluated molecular biomarkers of prognosis in HCC as putative biomarkers for LT success. Ultimately, we aimed to integrate clinical and molecular data in order to launch the basis for improved decision criteria, which after a prospective validation could be applied in a larger population. We first evaluated the long-term outcome of LT for HCC with domino LT, using livers from Familial Amyloidotic Polyneuropathy patients, with our original “double piggy-back technique”. Between 2001 and 2014, a total of 260 patients undergoing LT for HCC were analyzed from our dedicated database. We compared the outcomes of deceased-donor LT (146) and domino LT (114) using propensity score matching. Median follow-up was 34 months (1-152). Overall and disease-free 5-year survival rates for the total population were 58% and 56%, respectively. There were 177 (68%) patients within Milan Criteria and additionally 26 (10%) within UCSF Criteria. Patients older than 50 years were more likely to 174 receive a FAP liver (OR 1,94, CI 1.02-3.69). DLT patients had more major complications (23.7% versus 13.0%, p=0.025). Only patients undergoing DLT presented with “piggy-back syndrome” (7% versus 0%, p=0.001). After adjusting for potential confounders, DLT and cadaveric LT had a similar 5-year survival rate (59 % versus 44%, respectively, p=0.117). Thirteen patients (11.4%) evidenced FAP disease but not before 6 years after domino LT. This study concluded that domino LT for HCC is feasible and achieves equivalent results to cadaveric LT. The benefit of expanding the donor pool must be balanced against higher morbidity and a real risk of disease transmission. This was the first study to validate DLT in a population of cirrhotic patients with HCC. We then turned our attention to current clinical criteria for selection of patients for LT due to HCC. TTV (total tumor volume) has been proposed as a more accurate means of selecting patients for Liver Transplantation in this setting. TTV was already described in a population with a long waiting time on list (median of 8 months), achieving good results partially due to effect of the “test of time”. In this work, we aimed to analyze the role of TTV in a population with a short waiting time on list. Patients submitted to LT for HCC between September 1992 and February 2014 were studied. TTV, MC, UCSF and “Up to Seven” criteria were calculated both with pre-operative imaging exams and histological data. TTV was calculated as the sum of the volumes of all tumors based on its radius [(4/3)πr3, where r is the maximum radius of each HCC]. The cut-off value for TTV was selected to achieve the best sensitivity/specificity combination. The study population consisted of 231 out of patients. Median waiting time on list was 62.5 days. MC included 187 patients (81% of total), while TTV <115 cm3 included 214 (93% of the total population). Microvascular invasion (HR 2.601, C.I. 1.529-4.426), MC (HR 1.666, C.I. 0.990-2.804), UCSF Criteria (HR 2.995, C.I. 1.875-4.875), TTV < 115 cm3 (HR 2.898, C.I. 1.398-6.007) and “Up to Seven” Criteria (HR 2.139, C.I. 1.353-3.383) proved to be independent factors for prognosis for disease-free survival. This study concluded that TTV < 115 cm3 may be a useful tool to properly identify the best HCC candidates for liver transplantation in a population with a short waiting time on list. TTV allows the opportunity of liver transplantation for more patients, maintaining similar rates of tumor recurrence and patient survival. Finally, molecular markers of prognosis were studied. We combined bioinformatics data mining and a systematic review of the literature regarding prognosis after LR and LT. This approach revealed 16 genes whose gene expression levels had a potential prognostic impact in survival after LT or LR. To test these putative biomarkers in LT context, we studied a cohort of patients submitted to LT for HCC between September 1992 and February 2014. After applying inclusion and exclusion criteria, a total of 180 patients were obtained: 26 patients were first analyzed in a pilot set and 154 more where studied in a validation set. Archived FFPE tumoral tissue sections were submitted to total RNA extraction. Each extracted RNA was reverse‑transcribed to cDNA. Analysis of gene expression was done using quantitative realtime PCR. After correlating each gene expression to recurrence in a pilot set that was used to eliminate 11 of the original 16 candidate biomarkers, expression levels of CAPNS1, DPT, CLU, FBXW7 and SPRY2 were analyzed on the validation set and then applied to the total population. The expression of two genes, DPT and CLU, proved to be independently associated with decreased recurrence (DPT: OR 0.177, C.I. 0.064-0.489, p= 0.001; CLU: OR 0.385, C.I. 0.153-0.971, p=0.043), as well as increased OS and DFS (DPT: HR 0.507, C.I.0.310- 0.829, p=0.007; CLU: 0.443, C.I. 0.269-0.730, p=0.001). Strong DPT expression was associated with decreased microvascular invasion (OR 0.370, C.I. 0.140-0.979, p=0.045). Patients with a strong DPT expression experienced a 70% and 52.2% DFS at 5 and 10 years respectively. A strong expression of DPT allowed identification of patients in bad prognosis groups (beyond MC, TTV > 115 cm3, microvascular invasion, poorly-differentiated tumors), all experiencing 5- year survival over 70%. Patients with strong CLU expression exhibited a 5 and 10-year survival of 68.9% and 56.1% respectively. Again, a strong expression of CLU allowed identification of patients in bad prognosis groups (beyond MC, TTV > 115 cm3) experiencing 5-year survival over 60%. Finally, we analyzed a simple combined gene score including DPT and CLU. Strong expression of both genes resulted in 78% and 60% DFS at 5 and 10 years respectively, making the gene combination superior to DPT or CLU alone. Furthermore, association of the combined gene score with TTV measurement identified patients with good prognosis although having a TTV >115 cm3, as well as a group of patients with poor prognosis although having a TTV < 115 cm3. This is the first report that associates DPT expression and prognosis after LT for HCC and it is the first time this gene has proven to be an independent predictor of good prognosis. The results obtained by this gene combination are as good as the best previously published scores in the literature, with the advantage of simplicity. We conclude that the association of TTV < 115 cm3 and specimen expression of DPT and CLU could be, after proper validation, the basis to the development of a simple classification system that will allow more equity and accuracy in the process of selecting cirrhotic patients for LT due to HCC.
URI: http://hdl.handle.net/10362/31474
Designation: Doutoramento em Medicina, Especialidade em Cirurgia e Morfologia Humana
Appears in Collections:NMS-FCM - Teses de Doutoramento

Files in This Item:
File Description SizeFormat 
Marques Hugo Pinto TD 2018.pdf25,55 MBAdobe PDFView/Open    Request a copy


FacebookTwitterDeliciousLinkedInDiggGoogle BookmarksMySpace
Formato BibTex MendeleyEndnote 

Items in Repository are protected by copyright, with all rights reserved, unless otherwise indicated.