Pretransplant Therapy for Hepatocellular Carcinoma
Pretransplant Therapy for Hepatocellular Carcinoma
With the introduction of MC in 1996 followed by MELD prioritization of HCC recipients, LT for HCC has increased 6-fold over the last decade. Despite improved outcomes for recipients confined by these radiological size criteria, there is an increasing recognition that MELD prioritization of HCC recipients needs modification and should take into account the variable rates of tumor progression, waitlist dropout, and posttransplant recurrence. LRT, by inducing tumor necrosis, has emerged as an invaluable modality to mitigate tumor progression and waitlist dropout and potentially improve posttransplantation outcomes. We report herein the largest single-institution experience of LT for HCC in recipients receiving pretransplant LRT, focusing on the rate of cPR and its effect on posttransplant recurrence and survival.
One of our most striking findings was that achieving a cPR with total tumor necrosis was tantamount to cure. The 126 recipients (25% of total) with cPR developed only 3 recurrences and 1 cancer-specific death and had significantly superior overall and recurrence-free survival compared with patients with viable tumor. Although the rates of posttransplant HCC recurrence have been reported to be 0% to 35% after TACE and 0% to 13% after thermal radio frequency ablation ablation, our findings are consistent with the few studies that explicitly report on the recurrences in recipients with cPR. Mazzaferro et al reported no posttransplant recurrences in 33 of 60 RFA ablated lesions that demonstrated cPR, whereas Cucchetti et al reported an 8.6% 5-year recurrence rate in the 44 cPR recipients who underwent TACE. Most recently, Allard et al reported a 20% cPR rate after TACE in recipients undergoing LT, with a 6% 5-year recurrence in patients with cPR. Perhaps most importantly, the improvement in overall and recurrence-free survival was limited to LT recipients with at least 90% pathologic necrosis. The 15% recurrence rate in our recipients without cPR corroborates these findings and provides evidence that a partial response to LRT leaves behind viable tumor cells that may potentially disseminate in the peritransplant period, increasing the risk for a posttransplant recurrence.
Unlike the majority of studies evaluating a single LRT modality before LT, our experience afforded us the opportunity to evaluate the impact of multimodality treatments on the ability to achieve cPR. Of the 75 recipients receiving combination TACE and TA, 26 (35%) demonstrated cPR, compared with 29% receiving only TA and 19% receiving only TACE. Even after controlling for other significant predictors of cPR (recipient MELD, AFP, and cumulative tumor diameter), combination therapy was the strongest treatment-specific factor in predicting cPR. To our knowledge, only 1 other study has evaluated a combination of TACE and TA as a bridge to LT, reporting a 77% rate of cPR in 16 patients undergoing LT. Although the exact mechanisms for higher cPR rates remain unclear, it has been speculated that TACE before TA may eliminate the heat-sink effect allowing for creation of larger ablation zones, whereas TA before TACE creates a hyperemic rim surrounding the ablation area, which may then be more effectively targeted with TACE.
With cPR such a critical determinant of posttransplantation outcomes, we sought to evaluate the pretransplant factors that may predict cPR. Two of the most important predictive factors were related to the tumor response to LRT. The radiologic assessment subsequent to the final LRT and before LT was the strongest predictor, with a 2.4 and 5 times increased likelihood of achieving cPR for recipients demonstrating possible viable tumor or no viable tumor, compared with recipients with definite viable tumor. Our assessment was based on the presence or absence of persistent arterial enhancement in the treated lesion (s) or the development a new arterially enhancing lesion, slightly tailored from the widely accepted HCC mRECIST criteria to allow for a more simplified assessment aimed specifically to address the presence of viable carcinoma. Our results corroborate several recent studies demonstrating that the radiographic response to LRT accurately stratifies the risk of waitlist dropout and may be used to modify prioritization. A reduction in post-LRT AFP was also very important, with cPR 3 times less likely for HCC recipients who had an increase in AFP despite LRT (last AFP = max pre-LT AFP), consistent with recent studies demonstrating the importance of the AFP trend leading into LT.
Time spent on the waitlist for HCC recipients has emerged as a critical factor in predicting posttransplantation outcomes. Halazun et al recently analyzed national United Network for Organ Sharing data and identified significant disparities in waitlist times across United Network for Organ Sharing regions. Comparing long waiting time regions (regions 5 and 9, median waitlist time 7.6 months) with short waiting time regions (regions 3 and 10, median waitlist time 1.6 months), they found that although death on the waitlist was significantly lower in short waiting time region recipients (1.6% vs 8.4%), listing/transplantation in short waiting time regions was an independent predictor of poor patient survival (hazard ratio: 1.55, 95% CI: 1.38–1.74, P < 0.001). In addition, the excellent posttransplant outcomes after downstaging of beyond MC tumors have been attributed to a mandatory observation period after treatment to delineate tumor behavior. In our study, an increasing time interval from last LRT to transplantation significantly increased the likelihood of achieving cPR, which, we have shown, portends excellent posttransplant outcomes. Taken collectively, these data provide indisputable evidence that time on the waitlist is an excellent surrogate for tumor biology and may aid in the "weeding out" of aggressive tumors that are unlikely to be cured by transplantation.
(Enlarge Image)
Figure 4.
ROC curve of the multivariate model predicting the ability to achieve complete pathologic response to pretransplant locoregional therapy. ROC indicates receiver operating characteristic.
Finally, we found that a recipient's physiologic MELD score was a significant independent factor predicting cPR, with a 30% increased likelihood of cPR for every standard deviation decrease in the MELD. Although the MELD score is widely adopted as the gold standard in predicting waitlist mortality in all patients with ESLD, a relationship between MELD and the ability of LRT to achieve tumor necrosis has not previously been reported. Cucchetti et al reported that MELD was a significant predictor of waitlist dropout in HCC recipients; however, this increased risk of dropout was observed in 22 patients who did not receive any LRT, presumably due to more advanced liver disease precluding treatment (median MELD of 20). Although all recipients in our study had relatively well-compensated liver disease (median MELD 12), recipients demonstrating cPR had a slightly lower MELD than recipients without cPR. Because all of our patients received LRT, and recipients with cPR on average received fewer treatments, we propose that the effect of MELD on achieving tumor necrosis most likely reflects the increased aggressiveness of the individual treatments in recipients with lower MELD scores and potential MELD-driven variations in the host response to LRT. Further studies are needed to investigate the mechanisms by which the MELD score, as a marker of underlying liver function, may exert this effect.
A large body of evidence has shown that the current MELD allocation system not only overprioritizes HCC recipients but also fails to take into account the variable risks of tumor progression, waitlist dropout, and posttransplant recurrence within groups of patients confined by the same radiologic size criteria. In addition to tumor size and number, response to LRT, changes in AFP, and time on the waitlist are important surrogates of tumor biology, and consideration must be given to incorporate these factors into the HCC MELD allocation policy. Fortuitously, these same factors that predict the risk of waitlist dropout due to tumor progression also influence posttransplantation outcomes, aligning the goals of the ideal allocation policy to balance waitlist dropout due to tumor progression with favorable posttransplant outcomes.
Discussion
With the introduction of MC in 1996 followed by MELD prioritization of HCC recipients, LT for HCC has increased 6-fold over the last decade. Despite improved outcomes for recipients confined by these radiological size criteria, there is an increasing recognition that MELD prioritization of HCC recipients needs modification and should take into account the variable rates of tumor progression, waitlist dropout, and posttransplant recurrence. LRT, by inducing tumor necrosis, has emerged as an invaluable modality to mitigate tumor progression and waitlist dropout and potentially improve posttransplantation outcomes. We report herein the largest single-institution experience of LT for HCC in recipients receiving pretransplant LRT, focusing on the rate of cPR and its effect on posttransplant recurrence and survival.
One of our most striking findings was that achieving a cPR with total tumor necrosis was tantamount to cure. The 126 recipients (25% of total) with cPR developed only 3 recurrences and 1 cancer-specific death and had significantly superior overall and recurrence-free survival compared with patients with viable tumor. Although the rates of posttransplant HCC recurrence have been reported to be 0% to 35% after TACE and 0% to 13% after thermal radio frequency ablation ablation, our findings are consistent with the few studies that explicitly report on the recurrences in recipients with cPR. Mazzaferro et al reported no posttransplant recurrences in 33 of 60 RFA ablated lesions that demonstrated cPR, whereas Cucchetti et al reported an 8.6% 5-year recurrence rate in the 44 cPR recipients who underwent TACE. Most recently, Allard et al reported a 20% cPR rate after TACE in recipients undergoing LT, with a 6% 5-year recurrence in patients with cPR. Perhaps most importantly, the improvement in overall and recurrence-free survival was limited to LT recipients with at least 90% pathologic necrosis. The 15% recurrence rate in our recipients without cPR corroborates these findings and provides evidence that a partial response to LRT leaves behind viable tumor cells that may potentially disseminate in the peritransplant period, increasing the risk for a posttransplant recurrence.
Unlike the majority of studies evaluating a single LRT modality before LT, our experience afforded us the opportunity to evaluate the impact of multimodality treatments on the ability to achieve cPR. Of the 75 recipients receiving combination TACE and TA, 26 (35%) demonstrated cPR, compared with 29% receiving only TA and 19% receiving only TACE. Even after controlling for other significant predictors of cPR (recipient MELD, AFP, and cumulative tumor diameter), combination therapy was the strongest treatment-specific factor in predicting cPR. To our knowledge, only 1 other study has evaluated a combination of TACE and TA as a bridge to LT, reporting a 77% rate of cPR in 16 patients undergoing LT. Although the exact mechanisms for higher cPR rates remain unclear, it has been speculated that TACE before TA may eliminate the heat-sink effect allowing for creation of larger ablation zones, whereas TA before TACE creates a hyperemic rim surrounding the ablation area, which may then be more effectively targeted with TACE.
With cPR such a critical determinant of posttransplantation outcomes, we sought to evaluate the pretransplant factors that may predict cPR. Two of the most important predictive factors were related to the tumor response to LRT. The radiologic assessment subsequent to the final LRT and before LT was the strongest predictor, with a 2.4 and 5 times increased likelihood of achieving cPR for recipients demonstrating possible viable tumor or no viable tumor, compared with recipients with definite viable tumor. Our assessment was based on the presence or absence of persistent arterial enhancement in the treated lesion (s) or the development a new arterially enhancing lesion, slightly tailored from the widely accepted HCC mRECIST criteria to allow for a more simplified assessment aimed specifically to address the presence of viable carcinoma. Our results corroborate several recent studies demonstrating that the radiographic response to LRT accurately stratifies the risk of waitlist dropout and may be used to modify prioritization. A reduction in post-LRT AFP was also very important, with cPR 3 times less likely for HCC recipients who had an increase in AFP despite LRT (last AFP = max pre-LT AFP), consistent with recent studies demonstrating the importance of the AFP trend leading into LT.
Time spent on the waitlist for HCC recipients has emerged as a critical factor in predicting posttransplantation outcomes. Halazun et al recently analyzed national United Network for Organ Sharing data and identified significant disparities in waitlist times across United Network for Organ Sharing regions. Comparing long waiting time regions (regions 5 and 9, median waitlist time 7.6 months) with short waiting time regions (regions 3 and 10, median waitlist time 1.6 months), they found that although death on the waitlist was significantly lower in short waiting time region recipients (1.6% vs 8.4%), listing/transplantation in short waiting time regions was an independent predictor of poor patient survival (hazard ratio: 1.55, 95% CI: 1.38–1.74, P < 0.001). In addition, the excellent posttransplant outcomes after downstaging of beyond MC tumors have been attributed to a mandatory observation period after treatment to delineate tumor behavior. In our study, an increasing time interval from last LRT to transplantation significantly increased the likelihood of achieving cPR, which, we have shown, portends excellent posttransplant outcomes. Taken collectively, these data provide indisputable evidence that time on the waitlist is an excellent surrogate for tumor biology and may aid in the "weeding out" of aggressive tumors that are unlikely to be cured by transplantation.
(Enlarge Image)
Figure 4.
ROC curve of the multivariate model predicting the ability to achieve complete pathologic response to pretransplant locoregional therapy. ROC indicates receiver operating characteristic.
Finally, we found that a recipient's physiologic MELD score was a significant independent factor predicting cPR, with a 30% increased likelihood of cPR for every standard deviation decrease in the MELD. Although the MELD score is widely adopted as the gold standard in predicting waitlist mortality in all patients with ESLD, a relationship between MELD and the ability of LRT to achieve tumor necrosis has not previously been reported. Cucchetti et al reported that MELD was a significant predictor of waitlist dropout in HCC recipients; however, this increased risk of dropout was observed in 22 patients who did not receive any LRT, presumably due to more advanced liver disease precluding treatment (median MELD of 20). Although all recipients in our study had relatively well-compensated liver disease (median MELD 12), recipients demonstrating cPR had a slightly lower MELD than recipients without cPR. Because all of our patients received LRT, and recipients with cPR on average received fewer treatments, we propose that the effect of MELD on achieving tumor necrosis most likely reflects the increased aggressiveness of the individual treatments in recipients with lower MELD scores and potential MELD-driven variations in the host response to LRT. Further studies are needed to investigate the mechanisms by which the MELD score, as a marker of underlying liver function, may exert this effect.
A large body of evidence has shown that the current MELD allocation system not only overprioritizes HCC recipients but also fails to take into account the variable risks of tumor progression, waitlist dropout, and posttransplant recurrence within groups of patients confined by the same radiologic size criteria. In addition to tumor size and number, response to LRT, changes in AFP, and time on the waitlist are important surrogates of tumor biology, and consideration must be given to incorporate these factors into the HCC MELD allocation policy. Fortuitously, these same factors that predict the risk of waitlist dropout due to tumor progression also influence posttransplantation outcomes, aligning the goals of the ideal allocation policy to balance waitlist dropout due to tumor progression with favorable posttransplant outcomes.
Source...