- Top of page
Hepatocellular carcinoma (HCC) is the third leading cause of cancer-related death worldwide. This burden of disease is excepted to increase in the future, with the high prevalence of hepatitis B virus infections in Asia and sub-Saharan Africa, and the incidence of hepatitis C virus infections and alcoholic liver cirrhosis rising in developed regions.
The efficacy of liver transplantation for treatment of patients with HCC and cirrhosis was most notably described by Mazzaferro et al. in 1996 with the development of the Milan criteria. In a cohort of 48 patients with a single tumor 5 cm or less in diameter, or no more than three tumor nodules each 3 cm or less in diameter, liver transplantation achieved a 4-year overall survival rate of 92% and a disease-free survival rate of 85%. Despite being the most effective treatment, the shortage of available donor organs significantly reduces the efficacy of this treatment, with patients on waiting lists suffering significant disease progression.
Primary hepatic resection remains an accepted modality of treatment with 5-year overall survival rates of 55–71%.[5, 6] The continuous improvement in surgical technique and perioperative management has also reflected an improved survival outcomes with this treatment. However, recurrences are common, with almost 70% of patients developing intrahepatic or other disease recurrence within 5 years.[2, 7]
More recently, primary hepatic resection with curative intent followed by salvage liver transplantation (SLT) for those with disease recurrence has been promoted as a potential treatment modality. This strategy may potentially reduce disease progression for patients waiting for liver transplantation and reduce the number of transplants required. However, there remains concern over the potential for increased difficulty of transplantation following a prior resection and postoperative complications to negate the benefits of an SLT.
We propose to evaluate the outcomes of SLT for patients with recurrent HCC following initial treatment with primary hepatic resection. In this review, we seek to investigate using a systematic literature examination the morbidity, mortality, and survival outcomes of this therapeutic strategy.
- Top of page
Primary liver transplantation is recognized as the most effective treatment of primary HCC within the Milan criteria, but is limited by organ shortage. Efficacy of this treatment is affected by disease progression during prolonged waiting times. Primary hepatic resection is a widely adopted modality of treatment for primary HCC with reasonable long-term survival outcomes but is associated with high rates of disease recurrence. Poon et al. suggest a treatment strategy of primary hepatic resection as the treatment of patients with HCC within the Milan criteria, with SLT reserved for those with disease recurrence. This strategy may potentially reduce disease progression for patients waiting for liver transplantation and reduce the number of transplantations required. The pathological specimen obtained from a primary resection can also assist surgeons in identifying those patients at high risk of recurrence, who would most likely benefit from an SLT.[16, 37] The theoretical rate of patients eligible for SLT at recurrence has been reported to be as high as 60–80%.[8, 38] Although early clinical studies demonstrated the relative safety of this treatment strategy,[14, 20] there have been concerns about prior primary resection increasing the difficulty of SLT, negating potential outcome benefits.
Inclusion criteria for primary hepatic resection were generally consistent among studies. Initial resection was indicated in patients with good residual hepatic function, few tumor nodules (ideally solitary nodule), absence of intraoperative evidence of macrovascular invasion, absence of extrahepatic malignancy, and anatomically resectable disease. Early detection of recurrence is largely attributed to strict patient follow-up involving a combination of clinical history, examination, α-fetoprotein level, abdominal ultrasound, triple-phase CT scan, and magnetic resonance imaging at three to six monthly intervals for at least 12 months. Indications for SLT, as with primary transplantation, were consistent with disease within the Milan criteria. In addition, several SLT were performed on patients without disease recurrence, in the setting of hepatic decompensation[20, 24] and as a bridge transplantation.
This systematic review demonstrated reasonable rates of morbidity of the SLT strategy. Cumulative data from available studies in a recent systematic review by Maggs et al. suggest comparable rates of morbidity between primary transplantation and SLT. Of the studies included in our review, Moon et al. reported the largest series with results of 169 primary transplantations and 17 SLT. This study compared postoperative complications between primary transplantation and SLT, and did not demonstrate any significant differences between the rates of biliary (10.1 vs 17.6%, P = 0.401), bleeding (8.9 vs 11.8%, P = 0.658), vascular complications (1.8 vs 5.9%, P = 0.321), and the need for reoperation or retransplantation (4.1 vs 11.8%, P = 0.193). The length of hospital stay was also not significantly different between the two groups (37 vs 38 days, P = 0.566). Although operative time of salvage transplantation was increased when compared with primary transplantation in a number of studies, this difference was generally not significant.[28, 39, 40] Kaido et al. reported a retrospective analysis of living donor liver transplantations and demonstrated significantly increased operative time of SLT versus primary transplantation (941 min vs 763 min, P = 0.0024); however, this did not translate into differences in survival outcomes. Given the heterogeneous nature of studies included in this review and Maggs et al., it is difficult to draw further comparisons of morbidity results between primary transplantation and SLT without further studies with more consistent methodology.
The mortality rates associated with SLT following hepatic resection was significant (5%), but only three studies reporting mortality rates > 10%.[20, 32, 34] Shabahang et al. reported outcomes of primary hepatic resection versus primary liver transplantation and reported similar mortality rates (7 vs 7%). The mortality rate following primary liver transplantation was recorded in four of the studies (median 4%, range 2.1–7.0%, n = 744) and was similar to SLT.[20, 26, 29, 30] The rate of SLT following recurrence in our review was, however, significantly lower than the rates reported in theoretical studies.[8, 38]
This systematic review also demonstrates comparable overall and disease-free survival outcomes of the strategy of primary hepatic resection followed by SLT (median 62%, range 41–89%; median 67%, range 29–100%, respectively) when compared with primary liver transplantation (range 61–80%, range 58–89%, respectively).[20, 24, 29, 30, 32, 34, 35] In general, primary liver transplantation was associated with improved 5-year overall and disease-free survival, but these findings were only statistically significant in two studies,[20, 35] and disease-free survival but not overall survival was significantly improved with primary transplantation in two other studies.[30, 32]
The heterogeneous nature of currently available studies is recognized, and the heterogeneous cohort of patients may limit the ability for the results of this review to be extrapolated and compared against outcome data of other therapeutic modalities reported in the literature. The included studies either analyzed patients having previously undergone primary hepatic resection and subsequently SLT for recurrence, or retrospectively analyzed all patients receiving SLT to identify those who had received hepatic resection as treatment of primary disease. This variation in study design is reflected in data reporting. Studies employing the former study design[20, 21, 24, 25, 29, 31, 32] reported much higher median SLT rates of 41%, range 16–65%, when compared with median SLT rate 17%, range 7–36%, of purely retrospective studies.[22, 23, 26-28, 30, 33-35] It is recognized that the lack of randomized trials examining this treatment strategy also increases the potential risk of bias of the current literature.
Interestingly, Cucchetti et al. recently developed the Markov model to investigate the risk–benefit balance between primary liver transplantation and the treatment strategy discussed in this review. This model suggests that primary liver transplantation can produce improved survival outcomes when compared with primary hepatic resection and SLT if 5-year posttransplant survival remains higher than 60%. The balance between benefits and harm of SLT is clearly directly affected by the number of HCC candidates for transplantation and the expected waiting list time-to-transplant of local centers.
This review demonstrates that upfront primary hepatic resection is the treatment of choice in many centers with high incidence of HCC and significant organ shortage. In centers where all patients with HCC initially undergo hepatic resection, perhaps SLT should be viewed as one of many salvage treatment options. The comparison of SLT to other salvage treatment options is then more clinically relevant than comparisons with primary liver transplantations in such centers. Repeat hepatic resection is the only other potentially curative salvage therapy for recurrent HCC. A recent systematic review by our group on repeat hepatic resection as a salvage treatment option for recurrent HCC following primary resection demonstrates lower rates of morbidity and mortality, but worse disease-free and overall survival outcomes of repeat hepatic resection compared with SLT. The relationship between these two salvage treatment therapies is also mirrored in comparisons between hepatic resection and liver transplantation for primary disease.
The role of loco-regional therapy, in particular with the use of radiofrequency ablation (RFA), in recurrent HCC is still emerging. There is no evidence to support RFA as an alternative to SLT or repeat hepatic resection in patients with recurrent HCC, except in those unsuitable for operative management. Chan et al. reported a single-center retrospective series and demonstrated significantly poorer 5-year overall and disease-free survival outcomes with RFA compared with SLT or repeat hepatic resection (11% vs 50%, 48%). The role of RFA as neoadjuvant or adjuvant loco-regional therapy in relation to SLT is also unclear. Certainly for patients with disease exceeding the Milan criteria, RFA may be effective in downstaging the tumor; however, the limited evidence available does not currently support improved disease-free or overall survival in this setting.
Synthesis of available observational studies suggests that SLT following primary hepatic resection is a highly applicable treatment option with long-term survival outcomes and acceptable low rates of morbidity and mortality. Although no randomized studies between the two treatment strategies currently exist, the results of this review suggest that the tolerance and efficacy of these two treatment strategies may be comparable. The treatment strategy of primary hepatic resection followed by SLT may present an alternative to upfront liver transplantation with several potential benefits and is a clinical practice strategy that warrants further well-conducted randomized comparison study.