3.1. Safety, Efficacy, and Long-Term Outcomes
Y90 radioembolization’s role in HCC was first explored in the setting of portal vein thrombosis (PVT). It is estimated that approximately 1 in 3 HCC patients will develop PVT [21
] and historically, TACE in patients with PVT was relatively contraindicated out of concern for iatrogenically induced acute liver failure unless superselective or segmental embolization could be achieved. In 2004, Salem and colleagues reported on the safety of Y90 glass microspheres in unresectable HCC patients with PVT [22
]. In a cohort of 15 patients with PVT of one or both first order segmental portal vein branches who underwent Y90 radioembolization, radioembolization was well-tolerated with grade 1-2 bilirubin toxicity noted in only 5 patients. Importantly, this increase in post-treatment bilirubin was attributed to intrahepatic disease progression rather than treatment effect. No patients experienced liver failure, and for the first time, Y90 radioembolization was confirmed to be safe in HCC patients with PVT. Furthermore, it was found that due to the minimally embolic nature of Y90 glass microspheres, treatment with this form of radioembolization did not preclude future treatments with other arterial therapies such as TACE or hepatic artery infusion chemotherapy, paving the way for investigations into optimal treatment sequencing [22
]. In a similar two-center phase 2 study in 2008, the safety and efficacy of Y90 radioembolization using glass microspheres in 108 unresectable HCC patients with and without PVT of the branch or main portal vein was investigated [23
], building upon previous results in this patient population. For all patients, there were no cases of radiation-induced gastritis or pneumonitis. Post-treatment elevations in bilirubin and development of ascites were higher in the group with main PVT. Importantly, there was no increased risk of hepatic failure or encephalopathy between patients with main PVT versus those with branch or no PVT. Moreover, there were no significant treatment-related complications or treatment-related deaths. With respect to efficacy, partial responses were noted in 42.2% of patients using WHO criteria and in 70% of patients using EASL criteria. In 2013, treatment sequencing of HCC patients with PVT was further elucidated. In a cohort of 63 patients treated with Y90 radioembolization who experienced disease progression, investigators noted that 64% of patients were deemed ineligible for systemic treatment or clinical trials due to worsened Child-Pugh status (Child Pugh B or C) at time of progression [24
]. Additionally, survival was significantly shorter when comparing Child-Pugh A versus B groups (13.8 vs. 6.5 months). These findings led investigators to conclude that an adjuvant approach to sequencing, where systemic treatment is administered after Y90 radioembolization but before disease progression/declines in liver function, may be beneficial in this patient population [24
With the safety profile of Y90 radioembolization better understood and its efficacy coming to light, several studies began to examine long-term treatment outcomes across stages. In the first long term outcomes analysis for 291 HCC patients treated with Y90 radioembolization using glass microspheres, Salem et al. measured response rate, time to progression (TTP), and survival across stages [25
]. The investigators noted objective response rates (ORRs) of 42% and 57% based on WHO and EASL criteria, respectively. TTP for the entire patient cohort was 7.9 months and varied by Child-Pugh stage, as did survival outcomes. Patients with Child-Pugh A disease survived significantly longer than those with Child-Pugh B disease (17.2 vs. 7.7 months). Additionally, patients with Child-Pugh B disease and PVT had the worst outcomes with a median OS of 5.6 months. This study was the first to demonstrate the varying TTP and outcomes by Child-Pugh status. In a similar study of 108 intermediate-advanced HCC patients undergoing radioembolization with glass microspheres in Europe, TTP for the entire cohort was 10 months, and the median OS was 16.4 months [26
]. These outcomes compared favorably to the outcomes of similar patients treated with sorafenib monotherapy in the SHARP trial (TTP 5.5 months, median OS 10.7 months) [27
], setting the stage for combination studies discussed subsequently. In 2013, Mazzaferro and colleagues conducted the first phase 2 study examining the efficacy and long term outcomes of Y90 radioembolization in BCLC intermediate and advanced HCC [28
]. In 52 patients with a median follow up time of 36 months, the median TTP was 11 months and was not significantly different for patients with PVT versus those without PVT. Additionally, the median OS was 15 months and objective responses were noted in 40.4% of patients. These results were concordant with previous studies of radioembolization in intermediate-advanced HCC patients with or without PVT, leading researchers to conclude that intraarterial Y90 is a safe, well-tolerated, and efficacious treatment for these patients. More recently, a retrospective analysis by Gordon et al. sought to elucidate baseline patient characteristics and prognostic factors in unresectable HCC patients termed “Super Survivors”—those remaining alive >3 years after Y90 treatment [29
]. Sixty-seven “Super Survivors” were identified; surprisingly, they spanned BCLC stages A-D and 40% had multifocal disease. Upon further analysis, the common variable among these patients was an imaging response after treatment. Furthermore, patients undergoing segmental Y90 radioembolization were noted to have increased OS when compared to those undergoing lobar treatment. One possible confounder that may have influenced these results was crossover to alternate locoregional therapy after progression of disease. Nevertheless, imaging response to Y90 radioembolization treatment may be a favorable prognostic factor. In summary, over the past decade, Y90 radioembolization has demonstrated itself to be a safe and effective treatment for patients with unresectable HCC across clinical stages.
3.2. Comparisons to Existing Therapies
In recent years, there have been several studies comparing the efficacy of Y90 radioembolization to TACE and other locoregional therapies in the settings of definitive treatment and in downstaging as a bridge to transplant. In 2016, a systematic review and meta-analysis of 5 studies and 553 patients with unresectable HCC who underwent TACE or Y90 radioembolization was conducted [30
]. The analysis showed no significant survival differences for up to 4 years between the groups. Additionally, partial and complete response rates were similar, as were the complication profiles between the two treatments, although patients receiving TACE were noted to have increased post-treatment pain [30
]. These results differed from other meta-analyses around the same time, likely due to the retrospective nature of these analyses [31
]. In 2016, the results of the PREMIERE trial, a landmark phase 2 study comparing the TTP for 45 BCLC A and B patients randomized to TACE or Y90 were released [33
]. The investigators discovered a significantly longer median TTP for patients receiving Y90 radioembolization when compared to TACE (26 months vs. 6.8 months) although no differences in survival were noted, “suggesting local control is insufficient for survival improvement in cirrhotic patients with competing risks of death” [33
]. It should be noted that the trial was closed early due to slow accrual. Despite the lack of survival benefits, the researchers did note that the increased TTP and improved local control for patients treated with Y90 could decrease transplant list drop-out. Furthermore, lower rates of diarrhea and hypoalbuminemia, improved quality of life [34
], and outpatient delivery potentially make Y90 an attractive alternative to transplant-eligible HCC patients when compared to TACE. In the setting of downstaging for transplant eligibility, a comparative analysis of TACE versus Y90 demonstrated the superior performance of Y90 in downstaging HCC patients from United Network for Organ Sharing (UNOS) T3 to T2 [35
]. In this study of 86 patients, partial response rates were significantly higher in the Y90 patients, and downstaging was achieved in more Y90 patients when compared to TACE (58% versus 31%). By more efficiently downstaging patients from T3 to T2 and thus placing them within the confines of the Milan criteria, Y90 radioembolization may more quickly allow for UNOS priority status upgrade and quicker access to donor organs.
In addition to comparisons with TACE, the efficacy of Y90 radioembolization as a bridge to transplant has also been compared to stereotactic body radiation therapy (SBRT) and RFA. In a retrospective study of 60 HCC patients undergoing Y90, TACE, RFA, and SBRT, Mohammed et al. determined Y90 radioembolization to have the highest rates of pathologic complete response among the locoregional therapies (Y90/TACE/RFA/SBRT: 75%/41%/60%/28.5%) [36
]. Radiologic responses were also highest in the Y90 group at 33%, although this was not statistically significant. Finally, Grade 3 and 4 toxicities were significantly higher for patients receiving TACE and RFA. Although the choice of local therapy depends on numerous tumor and patient specific factors, Y90 has demonstrated itself to be an effective and tolerable local therapy for downstaging patients as a bridge to transplant [37
Several clinical trials have evaluated the role of Y90 radioembolization in combination with sorafenib. In 2014, Kulik and colleagues performed a phase 1 randomized trial to assess the safety and adverse event profile of Y90 radioembolization plus sorafenib compared to Y90 alone in 20 HCC patients awaiting a liver transplant [38
]. In the combination therapy group, all patients required dose reductions in sorafenib due to gastrointestinal and dermatologic side effects while 3 patients ultimately discontinued the drug. Importantly, in the combination group, peri-transplant (<30 day) biliary complications in four patients and acute cellular rejection (3/8 patients) were observed. These unexpected findings led the investigators to conclude that caution should be exercised when administering sorafenib plus Y90 radioembolization in the transplant setting. Additionally, survival rates were not significantly different between the two groups at 3 years (70% in monotherapy group, 72% in combination therapy group).
The relative efficacy of Y90 radioembolization versus sorafenib for advanced HCC patients was evaluated in the SARAH trial [39
]. This was a phase 3, randomized, controlled, open-label, multicenter trial that included 459 locally advanced HCC patients (BCLC C) or those previously treated with two unsuccessful rounds of TACE. Patients were randomized to sorafenib or Y90 radioembolization using resin microspheres and the primary end point was OS, with secondary endpoints including PFS, TTP, response rate, adverse events and quality of life (QOL). The median OS was not significantly different between the two treatment arms, with patients surviving 8.0 months in the radioembolization group and 9.9 months in the sorafenib group. Median PFS was similar between the two groups in both the intention-to-treat (ITT) population and in the per-protocol population. The objective response rate (ORR) was significantly higher in the Y90 radioembolization intention to treat population. Furthermore, higher rates of treatment related adverse events including fatigue, hematologic abnormalities, diarrhea, abdominal pain, and dermatologic reactions were noted in the sorafenib group.
More recently, a randomized, phase 3, open-label, multicenter trial comparing sorafenib to Y90 radioembolization using resin microspheres in 360 patients with locally advanced HCC patients with and without vascular invasion was conducted (SIRveNIB) [40
]. The primary endpoint was OS and secondary endpoint was PFS, tumor response rate, toxicity and QOL. Median OS was not statistically different between the two groups (8.8 versus 10.0 months in the Y90 radioembolization and sorafenib groups, respectively), although patients treated with Y90 radioembolization were noted to have higher tumor response rates and two-fold fewer adverse events. These results are similar to the SARAH study—both studies were negative and demonstrate no survival benefits for Y90 radioembolization when compared to sorafenib. Finally, it is worth noting that although these phase 3 trials measured injected radiation dose, actual dose delivered to the tumor was not measured, which has been shown to predict treatment response [41
]. Increased focus on Y90 radioembolization dosimetry moving forward may help to better elucidate treatment outcomes.
Preliminary results of a randomized, controlled, phase 2 trial comparing the safety and efficacy of sorafenib plus Y90 radioembolization using resin microspheres versus sorafenib alone in 529 patients with unresectable locally advanced HCC not amenable to TACE have been released (SORAMIC). The primary endpoint was OS and although median OS rates did not significantly differ between the two treatment groups in both the ITT and per-protocol population, possible survival benefits were noted in subgroups of younger patients (<65 years old), those with non-alcoholic cirrhosis, and those without cirrhosis at all [42
]. These initial results support further investigation into the optimal patient population to be treated with this combination therapy.